4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
43 compat_HTMLParseError
,
49 compat_ctypes_WINFUNCTYPE
,
50 compat_etree_fromstring
,
53 compat_html_entities_html5
,
66 compat_urllib_parse_urlencode
,
67 compat_urllib_parse_urlparse
,
68 compat_urllib_parse_urlunparse
,
69 compat_urllib_parse_quote
,
70 compat_urllib_parse_quote_plus
,
71 compat_urllib_parse_unquote_plus
,
72 compat_urllib_request
,
83 def register_socks_protocols():
84 # "Register" SOCKS protocols
85 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
86 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
87 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
88 if scheme
not in compat_urlparse
.uses_netloc
:
89 compat_urlparse
.uses_netloc
.append(scheme
)
92 # This is not clearly defined otherwise
93 compiled_regex_type
= type(re
.compile(''))
96 def random_user_agent():
97 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
1676 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
1680 'User-Agent': random_user_agent(),
1681 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
1682 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
1683 'Accept-Encoding': 'gzip, deflate',
1684 'Accept-Language': 'en-us,en;q=0.5',
1689 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
1693 NO_DEFAULT
= object()
1695 ENGLISH_MONTH_NAMES
= [
1696 'January', 'February', 'March', 'April', 'May', 'June',
1697 'July', 'August', 'September', 'October', 'November', 'December']
1700 'en': ENGLISH_MONTH_NAMES
,
1702 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
1703 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
1706 KNOWN_EXTENSIONS
= (
1707 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
1708 'flv', 'f4v', 'f4a', 'f4b',
1709 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
1710 'mkv', 'mka', 'mk3d',
1713 'asf', 'wmv', 'wma',
1719 'f4f', 'f4m', 'm3u8', 'smil')
1721 # needed for sanitizing filenames in restricted mode
1722 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
1723 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
1724 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
1748 '%Y/%m/%d %H:%M:%S',
1752 '%Y-%m-%d %H:%M:%S',
1753 '%Y-%m-%d %H:%M:%S.%f',
1754 '%Y-%m-%d %H:%M:%S:%f',
1757 '%Y-%m-%dT%H:%M:%SZ',
1758 '%Y-%m-%dT%H:%M:%S.%fZ',
1759 '%Y-%m-%dT%H:%M:%S.%f0Z',
1760 '%Y-%m-%dT%H:%M:%S',
1761 '%Y-%m-%dT%H:%M:%S.%f',
1763 '%b %d %Y at %H:%M',
1764 '%b %d %Y at %H:%M:%S',
1765 '%B %d %Y at %H:%M',
1766 '%B %d %Y at %H:%M:%S',
1770 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
1771 DATE_FORMATS_DAY_FIRST
.extend([
1777 '%d/%m/%Y %H:%M:%S',
1780 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
1781 DATE_FORMATS_MONTH_FIRST
.extend([
1786 '%m/%d/%Y %H:%M:%S',
1789 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
1790 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
1793 def preferredencoding():
1794 """Get preferred encoding.
1796 Returns the best encoding scheme for the system, based on
1797 locale.getpreferredencoding() and some further tweaks.
1800 pref = locale.getpreferredencoding()
1808 def write_json_file(obj, fn):
1809 """ Encode obj as JSON and write it to fn, atomically if possible """
1811 fn = encodeFilename(fn)
1812 if sys.version_info < (3, 0) and sys.platform != 'win32
':
1813 encoding = get_filesystem_encoding()
1814 # os.path.basename returns a bytes object, but NamedTemporaryFile
1815 # will fail if the filename contains non ascii characters unless we
1816 # use a unicode object
1817 path_basename = lambda f: os.path.basename(fn).decode(encoding)
1818 # the same for os.path.dirname
1819 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
1821 path_basename = os.path.basename
1822 path_dirname = os.path.dirname
1826 'prefix
': path_basename(fn) + '.',
1827 'dir': path_dirname(fn),
1831 # In Python 2.x, json.dump expects a bytestream.
1832 # In Python 3.x, it writes to a character stream
1833 if sys.version_info < (3, 0):
1838 'encoding
': 'utf
-8',
1841 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
1846 if sys.platform == 'win32
':
1847 # Need to remove existing file on Windows, else os.rename raises
1848 # WindowsError or FileExistsError.
1856 os.chmod(tf.name, 0o666 & ~mask)
1859 os.rename(tf.name, fn)
1868 if sys.version_info >= (2, 7):
1869 def find_xpath_attr(node, xpath, key, val=None):
1870 """ Find the xpath xpath[@key=val] """
1871 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
1872 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
1873 return node.find(expr)
1875 def find_xpath_attr(node, xpath, key, val=None):
1876 for f in node.findall(compat_xpath(xpath)):
1877 if key not in f.attrib:
1879 if val is None or f.attrib.get(key) == val:
1883 # On python2.6 the xml.etree.ElementTree.Element methods don't support
1884 # the namespace parameter
1887 def xpath_with_ns(path
, ns_map
):
1888 components
= [c
.split(':') for c
in path
.split('/')]
1890 for c
in components
:
1892 replaced
.append(c
[0])
1895 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
1896 return '/'.join(replaced
)
1899 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1900 def _find_xpath(xpath
):
1901 return node
.find(compat_xpath(xpath
))
1903 if isinstance(xpath
, (str, compat_str
)):
1904 n
= _find_xpath(xpath
)
1912 if default
is not NO_DEFAULT
:
1915 name
= xpath
if name
is None else name
1916 raise ExtractorError('Could not find XML element %s' % name
)
1922 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1923 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
1924 if n
is None or n
== default
:
1927 if default
is not NO_DEFAULT
:
1930 name
= xpath
if name
is None else name
1931 raise ExtractorError('Could not find XML element\'s text %s' % name
)
1937 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
1938 n
= find_xpath_attr(node
, xpath
, key
)
1940 if default
is not NO_DEFAULT
:
1943 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
1944 raise ExtractorError('Could not find XML attribute %s' % name
)
1947 return n
.attrib
[key
]
1950 def get_element_by_id(id, html
):
1951 """Return the content of the tag with the specified ID in the passed HTML document"""
1952 return get_element_by_attribute('id', id, html
)
1955 def get_element_by_class(class_name
, html
):
1956 """Return the content of the first tag with the specified class in the passed HTML document"""
1957 retval
= get_elements_by_class(class_name
, html
)
1958 return retval
[0] if retval
else None
1961 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
1962 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
1963 return retval
[0] if retval
else None
1966 def get_elements_by_class(class_name
, html
):
1967 """Return the content of all tags with the specified class in the passed HTML document as a list"""
1968 return get_elements_by_attribute(
1969 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
1970 html, escape_value=False)
1973 def get_elements_by_attribute(attribute, value, html, escape_value=True):
1974 """Return the content of the tag with the specified attribute in the passed HTML document"""
1976 value = re.escape(value) if escape_value else value
1979 for m in re.finditer(r'''(?xs)
1981 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^
']*'|
))*?
1983 (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^
"]*"|
='[^']*'|))*?
1987 ''' % (re.escape(attribute), value), html):
1988 res = m.group('content
')
1990 if res.startswith('"') or res.startswith("'"):
1993 retlist.append(unescapeHTML(res))
1998 class HTMLAttributeParser(compat_HTMLParser):
1999 """Trivial HTML parser to gather the attributes for a single element"""
2003 compat_HTMLParser.__init__(self)
2005 def handle_starttag(self, tag, attrs):
2006 self.attrs = dict(attrs)
2009 def extract_attributes(html_element):
2010 """Given a string for an HTML element such as
2012 a="foo" B="bar" c="&98;az" d=boz
2013 empty= noval entity="&"
2016 Decode and return a dictionary of attributes.
2018 'a
': 'foo
', 'b
': 'bar
', c: 'baz
', d: 'boz
',
2019 'empty
': '', 'noval
': None, 'entity
': '&',
2020 'sq
': '"', 'dq': '\''
2022 NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
2023 but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
2025 parser = HTMLAttributeParser()
2027 parser.feed(html_element)
2029 # Older Python may throw HTMLParseError in case of malformed HTML
2030 except compat_HTMLParseError:
2035 def clean_html(html):
2036 """Clean an HTML snippet into a readable string"""
2038 if html is None: # Convenience for sanitizing descriptions etc.
2042 html = html.replace('\n', ' ')
2043 html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
2044 html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
2046 html = re.sub('<.*?>', '', html)
2047 # Replace html entities
2048 html = unescapeHTML(html)
2052 def sanitize_open(filename, open_mode):
2053 """Try to open the given filename, and slightly tweak it if this fails.
2055 Attempts to open the given filename. If this fails, it tries to change
2056 the filename slightly, step by step, until it's either able to open it
2057 or it fails and raises a final exception, like the standard open()
2060 It returns the tuple (stream, definitive_file_name).
2064 if sys.platform == 'win32':
2066 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
2067 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
2068 stream = open(encodeFilename(filename), open_mode)
2069 return (stream, filename)
2070 except (IOError, OSError) as err:
2071 if err.errno in (errno.EACCES,):
2074 # In case of error, try to remove win32 forbidden chars
2075 alt_filename = sanitize_path(filename)
2076 if alt_filename == filename:
2079 # An exception here should be caught in the caller
2080 stream = open(encodeFilename(alt_filename), open_mode)
2081 return (stream, alt_filename)
2084 def timeconvert(timestr):
2085 """Convert RFC 2822 defined time string into system timestamp"""
2087 timetuple = email.utils.parsedate_tz(timestr)
2088 if timetuple is not None:
2089 timestamp = email.utils.mktime_tz(timetuple)
2093 def sanitize_filename(s, restricted=False, is_id=False):
2094 """Sanitizes a string so it could be used as part of a filename.
2095 If restricted is set, use a stricter subset of allowed characters.
2096 Set is_id if this is not an arbitrary string, but an ID that should be kept
2099 def replace_insane(char):
2100 if restricted and char in ACCENT_CHARS:
2101 return ACCENT_CHARS[char]
2102 elif not restricted and char == '\n':
2104 elif char == '?' or ord(char) < 32 or ord(char) == 127:
2107 return '' if restricted else '\''
2109 return '_
-' if restricted else ' -'
2110 elif char in '\\/|
*<>':
2112 if restricted and (char in '!&\'()[]{}$
;`^
,#' or char.isspace()):
2114 if restricted
and ord(char
) > 127:
2121 s
= re
.sub(r
'[0-9]+(?::[0-9]+)+', lambda m
: m
.group(0).replace(':', '_'), s
)
2122 result
= ''.join(map(replace_insane
, s
))
2124 while '__' in result
:
2125 result
= result
.replace('__', '_')
2126 result
= result
.strip('_')
2127 # Common case of "Foreign band name - English song title"
2128 if restricted
and result
.startswith('-_'):
2130 if result
.startswith('-'):
2131 result
= '_' + result
[len('-'):]
2132 result
= result
.lstrip('.')
2138 def sanitize_path(s
, force
=False):
2139 """Sanitizes and normalizes path on Windows"""
2140 if sys
.platform
== 'win32':
2142 drive_or_unc
, _
= os
.path
.splitdrive(s
)
2143 if sys
.version_info
< (2, 7) and not drive_or_unc
:
2144 drive_or_unc
, _
= os
.path
.splitunc(s
)
2150 norm_path
= os
.path
.normpath(remove_start(s
, drive_or_unc
)).split(os
.path
.sep
)
2154 path_part
if path_part
in ['.', '..'] else re
.sub(r
'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part
)
2155 for path_part
in norm_path
]
2157 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
2158 elif force
and s
[0] == os
.path
.sep
:
2159 sanitized_path
.insert(0, os
.path
.sep
)
2160 return os
.path
.join(*sanitized_path
)
2163 def sanitize_url(url
):
2164 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
2165 # the number of unwanted failures due to missing protocol
2166 if url
.startswith('//'):
2167 return 'http:%s' % url
2168 # Fix some common typos seen so far
2170 # https://github.com/ytdl-org/youtube-dl/issues/15649
2171 (r
'^httpss://', r
'https://'),
2172 # https://bx1.be/lives/direct-tv/
2173 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
2175 for mistake
, fixup
in COMMON_TYPOS
:
2176 if re
.match(mistake
, url
):
2177 return re
.sub(mistake
, fixup
, url
)
2181 def extract_basic_auth(url
):
2182 parts
= compat_urlparse
.urlsplit(url
)
2183 if parts
.username
is None:
2185 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
2186 parts
.hostname
if parts
.port
is None
2187 else '%s:%d' % (parts
.hostname
, parts
.port
))))
2188 auth_payload
= base64
.b64encode(
2189 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
2190 return url
, 'Basic ' + auth_payload
.decode('utf-8')
2193 def sanitized_Request(url
, *args
, **kwargs
):
2194 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
2195 if auth_header
is not None:
2196 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
2197 headers
['Authorization'] = auth_header
2198 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
2202 """Expand shell variables and ~"""
2203 return os
.path
.expandvars(compat_expanduser(s
))
2206 def orderedSet(iterable
):
2207 """ Remove all duplicates from the input iterable """
2215 def _htmlentity_transform(entity_with_semicolon
):
2216 """Transforms an HTML entity to a character."""
2217 entity
= entity_with_semicolon
[:-1]
2219 # Known non-numeric HTML entity
2220 if entity
in compat_html_entities
.name2codepoint
:
2221 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
2223 # TODO: HTML5 allows entities without a semicolon. For example,
2224 # 'Éric' should be decoded as 'Éric'.
2225 if entity_with_semicolon
in compat_html_entities_html5
:
2226 return compat_html_entities_html5
[entity_with_semicolon
]
2228 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
2229 if mobj
is not None:
2230 numstr
= mobj
.group(1)
2231 if numstr
.startswith('x'):
2233 numstr
= '0%s' % numstr
2236 # See https://github.com/ytdl-org/youtube-dl/issues/7518
2238 return compat_chr(int(numstr
, base
))
2242 # Unknown entity in name, return its literal representation
2243 return '&%s;' % entity
2246 def unescapeHTML(s
):
2249 assert type(s
) == compat_str
2252 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
2255 def escapeHTML(text
):
2258 .replace('&', '&')
2259 .replace('<', '<')
2260 .replace('>', '>')
2261 .replace('"', '"')
2262 .replace("'", ''')
2266 def process_communicate_or_kill(p
, *args
, **kwargs
):
2268 return p
.communicate(*args
, **kwargs
)
2269 except BaseException
: # Including KeyboardInterrupt
2275 def get_subprocess_encoding():
2276 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2277 # For subprocess calls, encode with locale encoding
2278 # Refer to http://stackoverflow.com/a/9951851/35070
2279 encoding
= preferredencoding()
2281 encoding
= sys
.getfilesystemencoding()
2282 if encoding
is None:
2287 def encodeFilename(s
, for_subprocess
=False):
2289 @param s The name of the file
2292 assert type(s
) == compat_str
2294 # Python 3 has a Unicode API
2295 if sys
.version_info
>= (3, 0):
2298 # Pass '' directly to use Unicode APIs on Windows 2000 and up
2299 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
2300 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
2301 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
2304 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
2305 if sys
.platform
.startswith('java'):
2308 return s
.encode(get_subprocess_encoding(), 'ignore')
2311 def decodeFilename(b
, for_subprocess
=False):
2313 if sys
.version_info
>= (3, 0):
2316 if not isinstance(b
, bytes):
2319 return b
.decode(get_subprocess_encoding(), 'ignore')
2322 def encodeArgument(s
):
2323 if not isinstance(s
, compat_str
):
2324 # Legacy code that uses byte strings
2325 # Uncomment the following line after fixing all post processors
2326 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
2327 s
= s
.decode('ascii')
2328 return encodeFilename(s
, True)
2331 def decodeArgument(b
):
2332 return decodeFilename(b
, True)
2335 def decodeOption(optval
):
2338 if isinstance(optval
, bytes):
2339 optval
= optval
.decode(preferredencoding())
2341 assert isinstance(optval
, compat_str
)
2345 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
2348 def timetuple_from_msec(msec
):
2349 secs
, msec
= divmod(msec
, 1000)
2350 mins
, secs
= divmod(secs
, 60)
2351 hrs
, mins
= divmod(mins
, 60)
2352 return _timetuple(hrs
, mins
, secs
, msec
)
2355 def formatSeconds(secs
, delim
=':', msec
=False):
2356 time
= timetuple_from_msec(secs
* 1000)
2358 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
2360 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
2362 ret
= '%d' % time
.seconds
2363 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
2366 def _ssl_load_windows_store_certs(ssl_context
, storename
):
2367 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
2369 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
2370 if encoding
== 'x509_asn' and (
2371 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
2372 except PermissionError
:
2376 ssl_context
.load_verify_locations(cadata
=cert
)
2377 except ssl
.SSLError
:
2381 def make_HTTPS_handler(params
, **kwargs
):
2382 opts_check_certificate
= not params
.get('nocheckcertificate')
2383 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
2384 context
.check_hostname
= opts_check_certificate
2385 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
2386 if opts_check_certificate
:
2388 context
.load_default_certs()
2389 # Work around the issue in load_default_certs when there are bad certificates. See:
2390 # https://github.com/yt-dlp/yt-dlp/issues/1060,
2391 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
2392 except ssl
.SSLError
:
2393 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
2394 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
2395 # Create a new context to discard any certificates that were already loaded
2396 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
2397 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
2398 for storename
in ('CA', 'ROOT'):
2399 _ssl_load_windows_store_certs(context
, storename
)
2400 context
.set_default_verify_paths()
2401 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
2404 def bug_reports_message(before
=';'):
2405 if ytdl_is_updateable():
2406 update_cmd
= 'type yt-dlp -U to update'
2408 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
2409 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
2410 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
2411 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
2413 before
= before
.rstrip()
2414 if not before
or before
.endswith(('.', '!', '?')):
2415 msg
= msg
[0].title() + msg
[1:]
2417 return (before
+ ' ' if before
else '') + msg
2420 class YoutubeDLError(Exception):
2421 """Base exception for YoutubeDL errors."""
2425 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
2426 if hasattr(ssl
, 'CertificateError'):
2427 network_exceptions
.append(ssl
.CertificateError
)
2428 network_exceptions
= tuple(network_exceptions
)
2431 class ExtractorError(YoutubeDLError
):
2432 """Error during info extraction."""
2434 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
2435 """ tb, if given, is the original traceback (so that it can be printed out).
2436 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
2438 if sys
.exc_info()[0] in network_exceptions
:
2443 self
.expected
= expected
2445 self
.video_id
= video_id
2447 self
.exc_info
= sys
.exc_info() # preserve original exception
2449 super(ExtractorError
, self
).__init
__(''.join((
2450 format_field(ie
, template
='[%s] '),
2451 format_field(video_id
, template
='%s: '),
2453 format_field(cause
, template
=' (caused by %r)'),
2454 '' if expected
else bug_reports_message())))
2456 def format_traceback(self
):
2457 if self
.traceback
is None:
2459 return ''.join(traceback
.format_tb(self
.traceback
))
2462 class UnsupportedError(ExtractorError
):
2463 def __init__(self
, url
):
2464 super(UnsupportedError
, self
).__init
__(
2465 'Unsupported URL: %s' % url
, expected
=True)
2469 class RegexNotFoundError(ExtractorError
):
2470 """Error when a regex didn't match"""
2474 class GeoRestrictedError(ExtractorError
):
2475 """Geographic restriction Error exception.
2477 This exception may be thrown when a video is not available from your
2478 geographic location due to geographic restrictions imposed by a website.
2481 def __init__(self
, msg
, countries
=None):
2482 super(GeoRestrictedError
, self
).__init
__(msg
, expected
=True)
2484 self
.countries
= countries
2487 class DownloadError(YoutubeDLError
):
2488 """Download Error exception.
2490 This exception may be thrown by FileDownloader objects if they are not
2491 configured to continue on errors. They will contain the appropriate
2495 def __init__(self
, msg
, exc_info
=None):
2496 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
2497 super(DownloadError
, self
).__init
__(msg
)
2498 self
.exc_info
= exc_info
2501 class EntryNotInPlaylist(YoutubeDLError
):
2502 """Entry not in playlist exception.
2504 This exception will be thrown by YoutubeDL when a requested entry
2505 is not found in the playlist info_dict
2510 class SameFileError(YoutubeDLError
):
2511 """Same File exception.
2513 This exception will be thrown by FileDownloader objects if they detect
2514 multiple files would have to be downloaded to the same file on disk.
2519 class PostProcessingError(YoutubeDLError
):
2520 """Post Processing exception.
2522 This exception may be raised by PostProcessor's .run() method to
2523 indicate an error in the postprocessing task.
2526 def __init__(self
, msg
):
2527 super(PostProcessingError
, self
).__init
__(msg
)
2531 class ExistingVideoReached(YoutubeDLError
):
2532 """ --max-downloads limit has been reached. """
2536 class RejectedVideoReached(YoutubeDLError
):
2537 """ --max-downloads limit has been reached. """
2541 class ThrottledDownload(YoutubeDLError
):
2542 """ Download speed below --throttled-rate. """
2546 class MaxDownloadsReached(YoutubeDLError
):
2547 """ --max-downloads limit has been reached. """
2551 class UnavailableVideoError(YoutubeDLError
):
2552 """Unavailable Format exception.
2554 This exception will be thrown when a video is requested
2555 in a format that is not available for that video.
2560 class ContentTooShortError(YoutubeDLError
):
2561 """Content Too Short exception.
2563 This exception may be raised by FileDownloader objects when a file they
2564 download is too small for what the server announced first, indicating
2565 the connection was probably interrupted.
2568 def __init__(self
, downloaded
, expected
):
2569 super(ContentTooShortError
, self
).__init
__(
2570 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
2573 self
.downloaded
= downloaded
2574 self
.expected
= expected
2577 class XAttrMetadataError(YoutubeDLError
):
2578 def __init__(self
, code
=None, msg
='Unknown error'):
2579 super(XAttrMetadataError
, self
).__init
__(msg
)
2583 # Parsing code and msg
2584 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
2585 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
2586 self
.reason
= 'NO_SPACE'
2587 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
2588 self
.reason
= 'VALUE_TOO_LONG'
2590 self
.reason
= 'NOT_SUPPORTED'
2593 class XAttrUnavailableError(YoutubeDLError
):
2597 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
2598 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
2599 # expected HTTP responses to meet HTTP/1.0 or later (see also
2600 # https://github.com/ytdl-org/youtube-dl/issues/6727)
2601 if sys
.version_info
< (3, 0):
2602 kwargs
['strict'] = True
2603 hc
= http_class(*args
, **compat_kwargs(kwargs
))
2604 source_address
= ydl_handler
._params
.get('source_address')
2606 if source_address
is not None:
2607 # This is to workaround _create_connection() from socket where it will try all
2608 # address data from getaddrinfo() including IPv6. This filters the result from
2609 # getaddrinfo() based on the source_address value.
2610 # This is based on the cpython socket.create_connection() function.
2611 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
2612 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
2613 host
, port
= address
2615 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
2616 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
2617 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
2618 if addrs
and not ip_addrs
:
2619 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
2621 "No remote IP%s addresses available for connect, can't use '%s' as source address"
2622 % (ip_version
, source_address
[0]))
2623 for res
in ip_addrs
:
2624 af
, socktype
, proto
, canonname
, sa
= res
2627 sock
= socket
.socket(af
, socktype
, proto
)
2628 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
2629 sock
.settimeout(timeout
)
2630 sock
.bind(source_address
)
2632 err
= None # Explicitly break reference cycle
2634 except socket
.error
as _
:
2636 if sock
is not None:
2641 raise socket
.error('getaddrinfo returns an empty list')
2642 if hasattr(hc
, '_create_connection'):
2643 hc
._create
_connection
= _create_connection
2644 sa
= (source_address
, 0)
2645 if hasattr(hc
, 'source_address'): # Python 2.7+
2646 hc
.source_address
= sa
2648 def _hc_connect(self
, *args
, **kwargs
):
2649 sock
= _create_connection(
2650 (self
.host
, self
.port
), self
.timeout
, sa
)
2652 self
.sock
= ssl
.wrap_socket(
2653 sock
, self
.key_file
, self
.cert_file
,
2654 ssl_version
=ssl
.PROTOCOL_TLSv1
)
2657 hc
.connect
= functools
.partial(_hc_connect
, hc
)
2662 def handle_youtubedl_headers(headers
):
2663 filtered_headers
= headers
2665 if 'Youtubedl-no-compression' in filtered_headers
:
2666 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
2667 del filtered_headers
['Youtubedl-no-compression']
2669 return filtered_headers
2672 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
2673 """Handler for HTTP requests and responses.
2675 This class, when installed with an OpenerDirector, automatically adds
2676 the standard headers to every HTTP request and handles gzipped and
2677 deflated responses from web servers. If compression is to be avoided in
2678 a particular request, the original request in the program code only has
2679 to include the HTTP header "Youtubedl-no-compression", which will be
2680 removed before making the real request.
2682 Part of this code was copied from:
2684 http://techknack.net/python-urllib2-handlers/
2686 Andrew Rowls, the author of that code, agreed to release it to the
2690 def __init__(self
, params
, *args
, **kwargs
):
2691 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
2692 self
._params
= params
2694 def http_open(self
, req
):
2695 conn_class
= compat_http_client
.HTTPConnection
2697 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2699 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2700 del req
.headers
['Ytdl-socks-proxy']
2702 return self
.do_open(functools
.partial(
2703 _create_http_connection
, self
, conn_class
, False),
2711 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
2713 return zlib
.decompress(data
)
2715 def http_request(self
, req
):
2716 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
2717 # always respected by websites, some tend to give out URLs with non percent-encoded
2718 # non-ASCII characters (see telemb.py, ard.py [#3412])
2719 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
2720 # To work around aforementioned issue we will replace request's original URL with
2721 # percent-encoded one
2722 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
2723 # the code of this workaround has been moved here from YoutubeDL.urlopen()
2724 url
= req
.get_full_url()
2725 url_escaped
= escape_url(url
)
2727 # Substitute URL if any change after escaping
2728 if url
!= url_escaped
:
2729 req
= update_Request(req
, url
=url_escaped
)
2731 for h
, v
in std_headers
.items():
2732 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
2733 # The dict keys are capitalized because of this bug by urllib
2734 if h
.capitalize() not in req
.headers
:
2735 req
.add_header(h
, v
)
2737 req
.headers
= handle_youtubedl_headers(req
.headers
)
2739 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
2740 # Python 2.6 is brain-dead when it comes to fragments
2741 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
2742 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
2746 def http_response(self
, req
, resp
):
2749 if resp
.headers
.get('Content-encoding', '') == 'gzip':
2750 content
= resp
.read()
2751 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
2753 uncompressed
= io
.BytesIO(gz
.read())
2754 except IOError as original_ioerror
:
2755 # There may be junk add the end of the file
2756 # See http://stackoverflow.com/q/4928560/35070 for details
2757 for i
in range(1, 1024):
2759 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
2760 uncompressed
= io
.BytesIO(gz
.read())
2765 raise original_ioerror
2766 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2767 resp
.msg
= old_resp
.msg
2768 del resp
.headers
['Content-encoding']
2770 if resp
.headers
.get('Content-encoding', '') == 'deflate':
2771 gz
= io
.BytesIO(self
.deflate(resp
.read()))
2772 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
2773 resp
.msg
= old_resp
.msg
2774 del resp
.headers
['Content-encoding']
2775 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
2776 # https://github.com/ytdl-org/youtube-dl/issues/6457).
2777 if 300 <= resp
.code
< 400:
2778 location
= resp
.headers
.get('Location')
2780 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
2781 if sys
.version_info
>= (3, 0):
2782 location
= location
.encode('iso-8859-1').decode('utf-8')
2784 location
= location
.decode('utf-8')
2785 location_escaped
= escape_url(location
)
2786 if location
!= location_escaped
:
2787 del resp
.headers
['Location']
2788 if sys
.version_info
< (3, 0):
2789 location_escaped
= location_escaped
.encode('utf-8')
2790 resp
.headers
['Location'] = location_escaped
2793 https_request
= http_request
2794 https_response
= http_response
2797 def make_socks_conn_class(base_class
, socks_proxy
):
2798 assert issubclass(base_class
, (
2799 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
2801 url_components
= compat_urlparse
.urlparse(socks_proxy
)
2802 if url_components
.scheme
.lower() == 'socks5':
2803 socks_type
= ProxyType
.SOCKS5
2804 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
2805 socks_type
= ProxyType
.SOCKS4
2806 elif url_components
.scheme
.lower() == 'socks4a':
2807 socks_type
= ProxyType
.SOCKS4A
2809 def unquote_if_non_empty(s
):
2812 return compat_urllib_parse_unquote_plus(s
)
2816 url_components
.hostname
, url_components
.port
or 1080,
2818 unquote_if_non_empty(url_components
.username
),
2819 unquote_if_non_empty(url_components
.password
),
2822 class SocksConnection(base_class
):
2824 self
.sock
= sockssocket()
2825 self
.sock
.setproxy(*proxy_args
)
2826 if type(self
.timeout
) in (int, float):
2827 self
.sock
.settimeout(self
.timeout
)
2828 self
.sock
.connect((self
.host
, self
.port
))
2830 if isinstance(self
, compat_http_client
.HTTPSConnection
):
2831 if hasattr(self
, '_context'): # Python > 2.6
2832 self
.sock
= self
._context
.wrap_socket(
2833 self
.sock
, server_hostname
=self
.host
)
2835 self
.sock
= ssl
.wrap_socket(self
.sock
)
2837 return SocksConnection
2840 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
2841 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
2842 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
2843 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
2844 self
._params
= params
2846 def https_open(self
, req
):
2848 conn_class
= self
._https
_conn
_class
2850 if hasattr(self
, '_context'): # python > 2.6
2851 kwargs
['context'] = self
._context
2852 if hasattr(self
, '_check_hostname'): # python 3.x
2853 kwargs
['check_hostname'] = self
._check
_hostname
2855 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
2857 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
2858 del req
.headers
['Ytdl-socks-proxy']
2860 return self
.do_open(functools
.partial(
2861 _create_http_connection
, self
, conn_class
, True),
2865 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
2867 See [1] for cookie file format.
2869 1. https://curl.haxx.se/docs/http-cookies.html
2871 _HTTPONLY_PREFIX
= '#HttpOnly_'
2873 _HEADER
= '''# Netscape HTTP Cookie File
2874 # This file is generated by yt-dlp. Do not edit.
2877 _CookieFileEntry
= collections
.namedtuple(
2879 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
2881 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2883 Save cookies to a file.
2885 Most of the code is taken from CPython 3.8 and slightly adapted
2886 to support cookie files with UTF-8 in both python 2 and 3.
2888 if filename
is None:
2889 if self
.filename
is not None:
2890 filename
= self
.filename
2892 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2894 # Store session cookies with `expires` set to 0 instead of an empty
2897 if cookie
.expires
is None:
2900 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
2901 f
.write(self
._HEADER
)
2904 if not ignore_discard
and cookie
.discard
:
2906 if not ignore_expires
and cookie
.is_expired(now
):
2912 if cookie
.domain
.startswith('.'):
2913 initial_dot
= 'TRUE'
2915 initial_dot
= 'FALSE'
2916 if cookie
.expires
is not None:
2917 expires
= compat_str(cookie
.expires
)
2920 if cookie
.value
is None:
2921 # cookies.txt regards 'Set-Cookie: foo' as a cookie
2922 # with no name, whereas http.cookiejar regards it as a
2923 # cookie with no value.
2928 value
= cookie
.value
2930 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
2931 secure
, expires
, name
, value
]) + '\n')
2933 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
2934 """Load cookies from a file."""
2935 if filename
is None:
2936 if self
.filename
is not None:
2937 filename
= self
.filename
2939 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
2941 def prepare_line(line
):
2942 if line
.startswith(self
._HTTPONLY
_PREFIX
):
2943 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
2944 # comments and empty lines are fine
2945 if line
.startswith('#') or not line
.strip():
2947 cookie_list
= line
.split('\t')
2948 if len(cookie_list
) != self
._ENTRY
_LEN
:
2949 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
2950 cookie
= self
._CookieFileEntry
(*cookie_list
)
2951 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
2952 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
2956 with io
.open(filename
, encoding
='utf-8') as f
:
2959 cf
.write(prepare_line(line
))
2960 except compat_cookiejar
.LoadError
as e
:
2962 'WARNING: skipping cookie file entry due to %s: %r\n'
2963 % (e
, line
), sys
.stderr
)
2966 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
2967 # Session cookies are denoted by either `expires` field set to
2968 # an empty string or 0. MozillaCookieJar only recognizes the former
2969 # (see [1]). So we need force the latter to be recognized as session
2970 # cookies on our own.
2971 # Session cookies may be important for cookies-based authentication,
2972 # e.g. usually, when user does not check 'Remember me' check box while
2973 # logging in on a site, some important cookies are stored as session
2974 # cookies so that not recognizing them will result in failed login.
2975 # 1. https://bugs.python.org/issue17164
2977 # Treat `expires=0` cookies as session cookies
2978 if cookie
.expires
== 0:
2979 cookie
.expires
= None
2980 cookie
.discard
= True
2983 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
2984 def __init__(self
, cookiejar
=None):
2985 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
2987 def http_response(self
, request
, response
):
2988 # Python 2 will choke on next HTTP request in row if there are non-ASCII
2989 # characters in Set-Cookie HTTP header of last response (see
2990 # https://github.com/ytdl-org/youtube-dl/issues/6769).
2991 # In order to at least prevent crashing we will percent encode Set-Cookie
2992 # header before HTTPCookieProcessor starts processing it.
2993 # if sys.version_info < (3, 0) and response.headers:
2994 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
2995 # set_cookie = response.headers.get(set_cookie_header)
2997 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
2998 # if set_cookie != set_cookie_escaped:
2999 # del response.headers[set_cookie_header]
3000 # response.headers[set_cookie_header] = set_cookie_escaped
3001 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
3003 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
3004 https_response
= http_response
3007 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
3008 """YoutubeDL redirect handler
3010 The code is based on HTTPRedirectHandler implementation from CPython [1].
3012 This redirect handler solves two issues:
3013 - ensures redirect URL is always unicode under python 2
3014 - introduces support for experimental HTTP response status code
3015 308 Permanent Redirect [2] used by some sites [3]
3017 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
3018 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
3019 3. https://github.com/ytdl-org/youtube-dl/issues/28768
3022 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
3024 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
3025 """Return a Request or None in response to a redirect.
3027 This is called by the http_error_30x methods when a
3028 redirection response is received. If a redirection should
3029 take place, return a new Request to allow http_error_30x to
3030 perform the redirect. Otherwise, raise HTTPError if no-one
3031 else should try to handle this url. Return None if you can't
3032 but another Handler might.
3034 m
= req
.get_method()
3035 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
3036 or code
in (301, 302, 303) and m
== "POST")):
3037 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
3038 # Strictly (according to RFC 2616), 301 or 302 in response to
3039 # a POST MUST NOT cause a redirection without confirmation
3040 # from the user (of urllib.request, in this case). In practice,
3041 # essentially all clients do redirect in this case, so we do
3044 # On python 2 urlh.geturl() may sometimes return redirect URL
3045 # as byte string instead of unicode. This workaround allows
3046 # to force it always return unicode.
3047 if sys
.version_info
[0] < 3:
3048 newurl
= compat_str(newurl
)
3050 # Be conciliant with URIs containing a space. This is mainly
3051 # redundant with the more complete encoding done in http_error_302(),
3052 # but it is kept for compatibility with other callers.
3053 newurl
= newurl
.replace(' ', '%20')
3055 CONTENT_HEADERS
= ("content-length", "content-type")
3056 # NB: don't use dict comprehension for python 2.6 compatibility
3057 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
3058 if k
.lower() not in CONTENT_HEADERS
)
3059 return compat_urllib_request
.Request(
3060 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
3064 def extract_timezone(date_str
):
3067 ^.{8,}? # >=8 char non-TZ prefix, if present
3068 (?P<tz>Z| # just the UTC Z, or
3069 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
3070 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
3071 [ ]? # optional space
3072 (?P<sign>\+|-) # +/-
3073 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
3077 timezone
= datetime
.timedelta()
3079 date_str
= date_str
[:-len(m
.group('tz'))]
3080 if not m
.group('sign'):
3081 timezone
= datetime
.timedelta()
3083 sign
= 1 if m
.group('sign') == '+' else -1
3084 timezone
= datetime
.timedelta(
3085 hours
=sign
* int(m
.group('hours')),
3086 minutes
=sign
* int(m
.group('minutes')))
3087 return timezone
, date_str
3090 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
3091 """ Return a UNIX timestamp from the given date """
3093 if date_str
is None:
3096 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
3098 if timezone
is None:
3099 timezone
, date_str
= extract_timezone(date_str
)
3102 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
3103 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
3104 return calendar
.timegm(dt
.timetuple())
3109 def date_formats(day_first
=True):
3110 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
3113 def unified_strdate(date_str
, day_first
=True):
3114 """Return a string with the date in the format YYYYMMDD"""
3116 if date_str
is None:
3120 date_str
= date_str
.replace(',', ' ')
3121 # Remove AM/PM + timezone
3122 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3123 _
, date_str
= extract_timezone(date_str
)
3125 for expression
in date_formats(day_first
):
3127 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
3130 if upload_date
is None:
3131 timetuple
= email
.utils
.parsedate_tz(date_str
)
3134 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
3137 if upload_date
is not None:
3138 return compat_str(upload_date
)
3141 def unified_timestamp(date_str
, day_first
=True):
3142 if date_str
is None:
3145 date_str
= re
.sub(r
'[,|]', '', date_str
)
3147 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
3148 timezone
, date_str
= extract_timezone(date_str
)
3150 # Remove AM/PM + timezone
3151 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
3153 # Remove unrecognized timezones from ISO 8601 alike timestamps
3154 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
3156 date_str
= date_str
[:-len(m
.group('tz'))]
3158 # Python only supports microseconds, so remove nanoseconds
3159 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
3161 date_str
= m
.group(1)
3163 for expression
in date_formats(day_first
):
3165 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
3166 return calendar
.timegm(dt
.timetuple())
3169 timetuple
= email
.utils
.parsedate_tz(date_str
)
3171 return calendar
.timegm(timetuple
) + pm_delta
* 3600
3174 def determine_ext(url
, default_ext
='unknown_video'):
3175 if url
is None or '.' not in url
:
3177 guess
= url
.partition('?')[0].rpartition('.')[2]
3178 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
3180 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
3181 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
3182 return guess
.rstrip('/')
3187 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
3188 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
3191 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
3193 Return a datetime object from a string in the format YYYYMMDD or
3194 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3196 format: string date format used to return datetime object from
3197 precision: round the time portion of a datetime object.
3198 auto|microsecond|second|minute|hour|day.
3199 auto: round to the unit provided in date_str (if applicable).
3201 auto_precision
= False
3202 if precision
== 'auto':
3203 auto_precision
= True
3204 precision
= 'microsecond'
3205 today
= datetime_round(datetime
.datetime
.now(), precision
)
3206 if date_str
in ('now', 'today'):
3208 if date_str
== 'yesterday':
3209 return today
- datetime
.timedelta(days
=1)
3211 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
3213 if match
is not None:
3214 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
3215 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
3216 unit
= match
.group('unit')
3217 if unit
== 'month' or unit
== 'year':
3218 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
3224 delta
= datetime
.timedelta(**{unit + 's': time}
)
3225 new_date
= start_time
+ delta
3227 return datetime_round(new_date
, unit
)
3230 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
3233 def date_from_str(date_str
, format
='%Y%m%d'):
3235 Return a datetime object from a string in the format YYYYMMDD or
3236 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
3238 format: string date format used to return datetime object from
3240 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
3243 def datetime_add_months(dt
, months
):
3244 """Increment/Decrement a datetime object by months."""
3245 month
= dt
.month
+ months
- 1
3246 year
= dt
.year
+ month
// 12
3247 month
= month
% 12 + 1
3248 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
3249 return dt
.replace(year
, month
, day
)
3252 def datetime_round(dt
, precision
='day'):
3254 Round a datetime object's time to a specific precision
3256 if precision
== 'microsecond':
3265 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
3266 timestamp
= calendar
.timegm(dt
.timetuple())
3267 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
3270 def hyphenate_date(date_str
):
3272 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
3273 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
3274 if match
is not None:
3275 return '-'.join(match
.groups())
3280 class DateRange(object):
3281 """Represents a time interval between two dates"""
3283 def __init__(self
, start
=None, end
=None):
3284 """start and end must be strings in the format accepted by date"""
3285 if start
is not None:
3286 self
.start
= date_from_str(start
)
3288 self
.start
= datetime
.datetime
.min.date()
3290 self
.end
= date_from_str(end
)
3292 self
.end
= datetime
.datetime
.max.date()
3293 if self
.start
> self
.end
:
3294 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
3298 """Returns a range that only contains the given day"""
3299 return cls(day
, day
)
3301 def __contains__(self
, date
):
3302 """Check if the date is in the range"""
3303 if not isinstance(date
, datetime
.date
):
3304 date
= date_from_str(date
)
3305 return self
.start
<= date
<= self
.end
3308 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
3311 def platform_name():
3312 """ Returns the platform name as a compat_str """
3313 res
= platform
.platform()
3314 if isinstance(res
, bytes):
3315 res
= res
.decode(preferredencoding())
3317 assert isinstance(res
, compat_str
)
3321 def get_windows_version():
3322 ''' Get Windows version. None if it's not running on Windows '''
3323 if compat_os_name
== 'nt':
3324 return version_tuple(platform
.win32_ver()[1])
3329 def _windows_write_string(s
, out
):
3330 """ Returns True if the string was written using special methods,
3331 False if it has yet to be written out."""
3332 # Adapted from http://stackoverflow.com/a/3259271/35070
3335 import ctypes
.wintypes
3343 fileno
= out
.fileno()
3344 except AttributeError:
3345 # If the output stream doesn't have a fileno, it's virtual
3347 except io
.UnsupportedOperation
:
3348 # Some strange Windows pseudo files?
3350 if fileno
not in WIN_OUTPUT_IDS
:
3353 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
3354 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
3355 ('GetStdHandle', ctypes
.windll
.kernel32
))
3356 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
3358 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
3359 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
3360 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
3361 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
3362 written
= ctypes
.wintypes
.DWORD(0)
3364 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
3365 FILE_TYPE_CHAR
= 0x0002
3366 FILE_TYPE_REMOTE
= 0x8000
3367 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
3368 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
3369 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
3370 ('GetConsoleMode', ctypes
.windll
.kernel32
))
3371 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
3373 def not_a_console(handle
):
3374 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
3376 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
3377 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
3379 if not_a_console(h
):
3382 def next_nonbmp_pos(s
):
3384 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
3385 except StopIteration:
3389 count
= min(next_nonbmp_pos(s
), 1024)
3391 ret
= WriteConsoleW(
3392 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
3394 raise OSError('Failed to write string')
3395 if not count
: # We just wrote a non-BMP character
3396 assert written
.value
== 2
3399 assert written
.value
> 0
3400 s
= s
[written
.value
:]
3404 def write_string(s
, out
=None, encoding
=None):
3407 assert type(s
) == compat_str
3409 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
3410 if _windows_write_string(s
, out
):
3413 if ('b' in getattr(out
, 'mode', '')
3414 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
3415 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
3417 elif hasattr(out
, 'buffer'):
3418 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
3419 byt
= s
.encode(enc
, 'ignore')
3420 out
.buffer.write(byt
)
3426 def bytes_to_intlist(bs
):
3429 if isinstance(bs
[0], int): # Python 3
3432 return [ord(c
) for c
in bs
]
3435 def intlist_to_bytes(xs
):
3438 return compat_struct_pack('%dB' % len(xs
), *xs
)
3441 # Cross-platform file locking
3442 if sys
.platform
== 'win32':
3443 import ctypes
.wintypes
3446 class OVERLAPPED(ctypes
.Structure
):
3448 ('Internal', ctypes
.wintypes
.LPVOID
),
3449 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
3450 ('Offset', ctypes
.wintypes
.DWORD
),
3451 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
3452 ('hEvent', ctypes
.wintypes
.HANDLE
),
3455 kernel32
= ctypes
.windll
.kernel32
3456 LockFileEx
= kernel32
.LockFileEx
3457 LockFileEx
.argtypes
= [
3458 ctypes
.wintypes
.HANDLE
, # hFile
3459 ctypes
.wintypes
.DWORD
, # dwFlags
3460 ctypes
.wintypes
.DWORD
, # dwReserved
3461 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3462 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3463 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3465 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
3466 UnlockFileEx
= kernel32
.UnlockFileEx
3467 UnlockFileEx
.argtypes
= [
3468 ctypes
.wintypes
.HANDLE
, # hFile
3469 ctypes
.wintypes
.DWORD
, # dwReserved
3470 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
3471 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
3472 ctypes
.POINTER(OVERLAPPED
) # Overlapped
3474 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
3475 whole_low
= 0xffffffff
3476 whole_high
= 0x7fffffff
3478 def _lock_file(f
, exclusive
):
3479 overlapped
= OVERLAPPED()
3480 overlapped
.Offset
= 0
3481 overlapped
.OffsetHigh
= 0
3482 overlapped
.hEvent
= 0
3483 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
3484 handle
= msvcrt
.get_osfhandle(f
.fileno())
3485 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
3486 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3487 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
3489 def _unlock_file(f
):
3490 assert f
._lock
_file
_overlapped
_p
3491 handle
= msvcrt
.get_osfhandle(f
.fileno())
3492 if not UnlockFileEx(handle
, 0,
3493 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
3494 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
3497 # Some platforms, such as Jython, is missing fcntl
3501 def _lock_file(f
, exclusive
):
3502 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
3504 def _unlock_file(f
):
3505 fcntl
.flock(f
, fcntl
.LOCK_UN
)
3507 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
3509 def _lock_file(f
, exclusive
):
3510 raise IOError(UNSUPPORTED_MSG
)
3512 def _unlock_file(f
):
3513 raise IOError(UNSUPPORTED_MSG
)
3516 class locked_file(object):
3517 def __init__(self
, filename
, mode
, encoding
=None):
3518 assert mode
in ['r', 'a', 'w']
3519 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
3522 def __enter__(self
):
3523 exclusive
= self
.mode
!= 'r'
3525 _lock_file(self
.f
, exclusive
)
3531 def __exit__(self
, etype
, value
, traceback
):
3533 _unlock_file(self
.f
)
3540 def write(self
, *args
):
3541 return self
.f
.write(*args
)
3543 def read(self
, *args
):
3544 return self
.f
.read(*args
)
3547 def get_filesystem_encoding():
3548 encoding
= sys
.getfilesystemencoding()
3549 return encoding
if encoding
is not None else 'utf-8'
3552 def shell_quote(args
):
3554 encoding
= get_filesystem_encoding()
3556 if isinstance(a
, bytes):
3557 # We may get a filename encoded with 'encodeFilename'
3558 a
= a
.decode(encoding
)
3559 quoted_args
.append(compat_shlex_quote(a
))
3560 return ' '.join(quoted_args
)
3563 def smuggle_url(url
, data
):
3564 """ Pass additional data in a URL for internal use. """
3566 url
, idata
= unsmuggle_url(url
, {})
3568 sdata
= compat_urllib_parse_urlencode(
3569 {'__youtubedl_smuggle': json.dumps(data)}
)
3570 return url
+ '#' + sdata
3573 def unsmuggle_url(smug_url
, default
=None):
3574 if '#__youtubedl_smuggle' not in smug_url
:
3575 return smug_url
, default
3576 url
, _
, sdata
= smug_url
.rpartition('#')
3577 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
3578 data
= json
.loads(jsond
)
3582 def format_bytes(bytes):
3585 if type(bytes) is str:
3586 bytes = float(bytes)
3590 exponent
= int(math
.log(bytes, 1024.0))
3591 suffix
= ['B', 'KiB', 'MiB', 'GiB', 'TiB', 'PiB', 'EiB', 'ZiB', 'YiB'][exponent
]
3592 converted
= float(bytes) / float(1024 ** exponent
)
3593 return '%.2f%s' % (converted
, suffix
)
3596 def lookup_unit_table(unit_table
, s
):
3597 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
3599 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
3602 num_str
= m
.group('num').replace(',', '.')
3603 mult
= unit_table
[m
.group('unit')]
3604 return int(float(num_str
) * mult
)
3607 def parse_filesize(s
):
3611 # The lower-case forms are of course incorrect and unofficial,
3612 # but we support those too
3629 'megabytes': 1000 ** 2,
3630 'mebibytes': 1024 ** 2,
3636 'gigabytes': 1000 ** 3,
3637 'gibibytes': 1024 ** 3,
3643 'terabytes': 1000 ** 4,
3644 'tebibytes': 1024 ** 4,
3650 'petabytes': 1000 ** 5,
3651 'pebibytes': 1024 ** 5,
3657 'exabytes': 1000 ** 6,
3658 'exbibytes': 1024 ** 6,
3664 'zettabytes': 1000 ** 7,
3665 'zebibytes': 1024 ** 7,
3671 'yottabytes': 1000 ** 8,
3672 'yobibytes': 1024 ** 8,
3675 return lookup_unit_table(_UNIT_TABLE
, s
)
3684 if re
.match(r
'^[\d,.]+$', s
):
3685 return str_to_int(s
)
3696 return lookup_unit_table(_UNIT_TABLE
, s
)
3699 def parse_resolution(s
):
3703 mobj
= re
.search(r
'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s
)
3706 'width': int(mobj
.group('w')),
3707 'height': int(mobj
.group('h')),
3710 mobj
= re
.search(r
'\b(\d+)[pPiI]\b', s
)
3712 return {'height': int(mobj.group(1))}
3714 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
3716 return {'height': int(mobj.group(1)) * 540}
3721 def parse_bitrate(s
):
3722 if not isinstance(s
, compat_str
):
3724 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
3726 return int(mobj
.group(1))
3729 def month_by_name(name
, lang
='en'):
3730 """ Return the number of a month by (locale-independently) English name """
3732 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
3735 return month_names
.index(name
) + 1
3740 def month_by_abbreviation(abbrev
):
3741 """ Return the number of a month by (locale-independently) English
3745 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
3750 def fix_xml_ampersands(xml_str
):
3751 """Replace all the '&' by '&' in XML"""
3753 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
3758 def setproctitle(title
):
3759 assert isinstance(title
, compat_str
)
3761 # ctypes in Jython is not complete
3762 # http://bugs.jython.org/issue2148
3763 if sys
.platform
.startswith('java'):
3767 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
3771 # LoadLibrary in Windows Python 2.7.13 only expects
3772 # a bytestring, but since unicode_literals turns
3773 # every string into a unicode string, it fails.
3775 title_bytes
= title
.encode('utf-8')
3776 buf
= ctypes
.create_string_buffer(len(title_bytes
))
3777 buf
.value
= title_bytes
3779 libc
.prctl(15, buf
, 0, 0, 0)
3780 except AttributeError:
3781 return # Strange libc, just skip this
3784 def remove_start(s
, start
):
3785 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
3788 def remove_end(s
, end
):
3789 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
3792 def remove_quotes(s
):
3793 if s
is None or len(s
) < 2:
3795 for quote
in ('"', "'", ):
3796 if s
[0] == quote
and s
[-1] == quote
:
3801 def get_domain(url
):
3802 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
3803 return domain
.group('domain') if domain
else None
3806 def url_basename(url
):
3807 path
= compat_urlparse
.urlparse(url
).path
3808 return path
.strip('/').split('/')[-1]
3812 return re
.match(r
'https?://[^?#&]+/', url
).group()
3815 def urljoin(base
, path
):
3816 if isinstance(path
, bytes):
3817 path
= path
.decode('utf-8')
3818 if not isinstance(path
, compat_str
) or not path
:
3820 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
3822 if isinstance(base
, bytes):
3823 base
= base
.decode('utf-8')
3824 if not isinstance(base
, compat_str
) or not re
.match(
3825 r
'^(?:https?:)?//', base
):
3827 return compat_urlparse
.urljoin(base
, path
)
3830 class HEADRequest(compat_urllib_request
.Request
):
3831 def get_method(self
):
3835 class PUTRequest(compat_urllib_request
.Request
):
3836 def get_method(self
):
3840 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
3843 v
= getattr(v
, get_attr
, None)
3849 return int(v
) * invscale
// scale
3850 except (ValueError, TypeError):
3854 def str_or_none(v
, default
=None):
3855 return default
if v
is None else compat_str(v
)
3858 def str_to_int(int_str
):
3859 """ A more relaxed version of int_or_none """
3860 if isinstance(int_str
, compat_integer_types
):
3862 elif isinstance(int_str
, compat_str
):
3863 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
3864 return int_or_none(int_str
)
3867 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
3871 return float(v
) * invscale
/ scale
3872 except (ValueError, TypeError):
3876 def bool_or_none(v
, default
=None):
3877 return v
if isinstance(v
, bool) else default
3880 def strip_or_none(v
, default
=None):
3881 return v
.strip() if isinstance(v
, compat_str
) else default
3884 def url_or_none(url
):
3885 if not url
or not isinstance(url
, compat_str
):
3888 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
3891 def strftime_or_none(timestamp
, date_format
, default
=None):
3892 datetime_object
= None
3894 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
3895 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
3896 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
3897 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
3898 return datetime_object
.strftime(date_format
)
3899 except (ValueError, TypeError, AttributeError):
3903 def parse_duration(s
):
3904 if not isinstance(s
, compat_basestring
):
3909 days
, hours
, mins
, secs
, ms
= [None] * 5
3910 m
= re
.match(r
'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s
)
3912 days
, hours
, mins
, secs
, ms
= m
.groups()
3917 [0-9]+\s*y(?:ears?)?\s*
3920 [0-9]+\s*m(?:onths?)?\s*
3923 [0-9]+\s*w(?:eeks?)?\s*
3926 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
3930 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
3933 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
3936 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
3939 days
, hours
, mins
, secs
, ms
= m
.groups()
3941 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
3943 hours
, mins
= m
.groups()
3949 duration
+= float(secs
)
3951 duration
+= float(mins
) * 60
3953 duration
+= float(hours
) * 60 * 60
3955 duration
+= float(days
) * 24 * 60 * 60
3957 duration
+= float(ms
)
3961 def prepend_extension(filename
, ext
, expected_real_ext
=None):
3962 name
, real_ext
= os
.path
.splitext(filename
)
3964 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
3965 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
3966 else '{0}.{1}'.format(filename
, ext
))
3969 def replace_extension(filename
, ext
, expected_real_ext
=None):
3970 name
, real_ext
= os
.path
.splitext(filename
)
3971 return '{0}.{1}'.format(
3972 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
3976 def check_executable(exe
, args
=[]):
3977 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
3978 args can be a list of arguments for a short output (like -version) """
3980 process_communicate_or_kill(subprocess
.Popen(
3981 [exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
))
3987 def get_exe_version(exe
, args
=['--version'],
3988 version_re
=None, unrecognized
='present'):
3989 """ Returns the version of the specified executable,
3990 or False if the executable is not present """
3992 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
3993 # SIGTTOU if yt-dlp is run in the background.
3994 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
3995 out
, _
= process_communicate_or_kill(subprocess
.Popen(
3996 [encodeArgument(exe
)] + args
,
3997 stdin
=subprocess
.PIPE
,
3998 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
))
4001 if isinstance(out
, bytes): # Python 2.x
4002 out
= out
.decode('ascii', 'ignore')
4003 return detect_exe_version(out
, version_re
, unrecognized
)
4006 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
4007 assert isinstance(output
, compat_str
)
4008 if version_re
is None:
4009 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
4010 m
= re
.search(version_re
, output
)
4017 class LazyList(collections
.abc
.Sequence
):
4018 ''' Lazy immutable list from an iterable
4019 Note that slices of a LazyList are lists and not LazyList'''
4021 class IndexError(IndexError):
4024 def __init__(self
, iterable
):
4025 self
.__iterable
= iter(iterable
)
4027 self
.__reversed
= False
4031 # We need to consume the entire iterable to iterate in reverse
4032 yield from self
.exhaust()
4034 yield from self
.__cache
4035 for item
in self
.__iterable
:
4036 self
.__cache
.append(item
)
4039 def __exhaust(self
):
4040 self
.__cache
.extend(self
.__iterable
)
4044 ''' Evaluate the entire iterable '''
4045 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
4048 def __reverse_index(x
):
4049 return None if x
is None else -(x
+ 1)
4051 def __getitem__(self
, idx
):
4052 if isinstance(idx
, slice):
4054 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
4055 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
4056 elif isinstance(idx
, int):
4058 idx
= self
.__reverse
_index
(idx
)
4059 start
, stop
, step
= idx
, idx
, 0
4061 raise TypeError('indices must be integers or slices')
4062 if ((start
or 0) < 0 or (stop
or 0) < 0
4063 or (start
is None and step
< 0)
4064 or (stop
is None and step
> 0)):
4065 # We need to consume the entire iterable to be able to slice from the end
4066 # Obviously, never use this with infinite iterables
4069 return self
.__cache
[idx
]
4070 except IndexError as e
:
4071 raise self
.IndexError(e
) from e
4072 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
4074 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
4076 return self
.__cache
[idx
]
4077 except IndexError as e
:
4078 raise self
.IndexError(e
) from e
4082 self
[-1] if self
.__reversed
else self
[0]
4083 except self
.IndexError:
4089 return len(self
.__cache
)
4092 self
.__reversed
= not self
.__reversed
4096 # repr and str should mimic a list. So we exhaust the iterable
4097 return repr(self
.exhaust())
4100 return repr(self
.exhaust())
4105 # This is only useful for tests
4106 return len(self
.getslice())
4108 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
4109 self
._pagefunc
= pagefunc
4110 self
._pagesize
= pagesize
4111 self
._use
_cache
= use_cache
4114 def getpage(self
, pagenum
):
4115 page_results
= self
._cache
.get(pagenum
) or list(self
._pagefunc
(pagenum
))
4117 self
._cache
[pagenum
] = page_results
4120 def getslice(self
, start
=0, end
=None):
4121 return list(self
._getslice
(start
, end
))
4123 def _getslice(self
, start
, end
):
4124 raise NotImplementedError('This method must be implemented by subclasses')
4126 def __getitem__(self
, idx
):
4127 # NOTE: cache must be enabled if this is used
4128 if not isinstance(idx
, int) or idx
< 0:
4129 raise TypeError('indices must be non-negative integers')
4130 entries
= self
.getslice(idx
, idx
+ 1)
4131 return entries
[0] if entries
else None
4134 class OnDemandPagedList(PagedList
):
4135 def _getslice(self
, start
, end
):
4136 for pagenum
in itertools
.count(start
// self
._pagesize
):
4137 firstid
= pagenum
* self
._pagesize
4138 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
4139 if start
>= nextfirstid
:
4143 start
% self
._pagesize
4144 if firstid
<= start
< nextfirstid
4147 ((end
- 1) % self
._pagesize
) + 1
4148 if (end
is not None and firstid
<= end
<= nextfirstid
)
4151 page_results
= self
.getpage(pagenum
)
4152 if startv
!= 0 or endv
is not None:
4153 page_results
= page_results
[startv
:endv
]
4154 yield from page_results
4156 # A little optimization - if current page is not "full", ie. does
4157 # not contain page_size videos then we can assume that this page
4158 # is the last one - there are no more ids on further pages -
4159 # i.e. no need to query again.
4160 if len(page_results
) + startv
< self
._pagesize
:
4163 # If we got the whole page, but the next page is not interesting,
4164 # break out early as well
4165 if end
== nextfirstid
:
4169 class InAdvancePagedList(PagedList
):
4170 def __init__(self
, pagefunc
, pagecount
, pagesize
):
4171 self
._pagecount
= pagecount
4172 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
4174 def _getslice(self
, start
, end
):
4175 start_page
= start
// self
._pagesize
4177 self
._pagecount
if end
is None else (end
// self
._pagesize
+ 1))
4178 skip_elems
= start
- start_page
* self
._pagesize
4179 only_more
= None if end
is None else end
- start
4180 for pagenum
in range(start_page
, end_page
):
4181 page_results
= self
.getpage(pagenum
)
4183 page_results
= page_results
[skip_elems
:]
4185 if only_more
is not None:
4186 if len(page_results
) < only_more
:
4187 only_more
-= len(page_results
)
4189 yield from page_results
[:only_more
]
4191 yield from page_results
4194 def uppercase_escape(s
):
4195 unicode_escape
= codecs
.getdecoder('unicode_escape')
4197 r
'\\U[0-9a-fA-F]{8}',
4198 lambda m
: unicode_escape(m
.group(0))[0],
4202 def lowercase_escape(s
):
4203 unicode_escape
= codecs
.getdecoder('unicode_escape')
4205 r
'\\u[0-9a-fA-F]{4}',
4206 lambda m
: unicode_escape(m
.group(0))[0],
4210 def escape_rfc3986(s
):
4211 """Escape non-ASCII characters as suggested by RFC 3986"""
4212 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
4213 s
= s
.encode('utf-8')
4214 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
4217 def escape_url(url
):
4218 """Escape URL as suggested by RFC 3986"""
4219 url_parsed
= compat_urllib_parse_urlparse(url
)
4220 return url_parsed
._replace
(
4221 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
4222 path
=escape_rfc3986(url_parsed
.path
),
4223 params
=escape_rfc3986(url_parsed
.params
),
4224 query
=escape_rfc3986(url_parsed
.query
),
4225 fragment
=escape_rfc3986(url_parsed
.fragment
)
4230 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
4233 def read_batch_urls(batch_fd
):
4235 if not isinstance(url
, compat_str
):
4236 url
= url
.decode('utf-8', 'replace')
4237 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
4238 for bom
in BOM_UTF8
:
4239 if url
.startswith(bom
):
4240 url
= url
[len(bom
):]
4242 if not url
or url
.startswith(('#', ';', ']')):
4244 # "#" cannot be stripped out since it is part of the URI
4245 # However, it can be safely stipped out if follwing a whitespace
4246 return re
.split(r
'\s#', url
, 1)[0].rstrip()
4248 with contextlib
.closing(batch_fd
) as fd
:
4249 return [url
for url
in map(fixup
, fd
) if url
]
4252 def urlencode_postdata(*args
, **kargs
):
4253 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
4256 def update_url_query(url
, query
):
4259 parsed_url
= compat_urlparse
.urlparse(url
)
4260 qs
= compat_parse_qs(parsed_url
.query
)
4262 return compat_urlparse
.urlunparse(parsed_url
._replace
(
4263 query
=compat_urllib_parse_urlencode(qs
, True)))
4266 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
4267 req_headers
= req
.headers
.copy()
4268 req_headers
.update(headers
)
4269 req_data
= data
or req
.data
4270 req_url
= update_url_query(url
or req
.get_full_url(), query
)
4271 req_get_method
= req
.get_method()
4272 if req_get_method
== 'HEAD':
4273 req_type
= HEADRequest
4274 elif req_get_method
== 'PUT':
4275 req_type
= PUTRequest
4277 req_type
= compat_urllib_request
.Request
4279 req_url
, data
=req_data
, headers
=req_headers
,
4280 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
4281 if hasattr(req
, 'timeout'):
4282 new_req
.timeout
= req
.timeout
4286 def _multipart_encode_impl(data
, boundary
):
4287 content_type
= 'multipart/form-data; boundary=%s' % boundary
4290 for k
, v
in data
.items():
4291 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
4292 if isinstance(k
, compat_str
):
4293 k
= k
.encode('utf-8')
4294 if isinstance(v
, compat_str
):
4295 v
= v
.encode('utf-8')
4296 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
4297 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
4298 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
4299 if boundary
.encode('ascii') in content
:
4300 raise ValueError('Boundary overlaps with data')
4303 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
4305 return out
, content_type
4308 def multipart_encode(data
, boundary
=None):
4310 Encode a dict to RFC 7578-compliant form-data
4313 A dict where keys and values can be either Unicode or bytes-like
4316 If specified a Unicode object, it's used as the boundary. Otherwise
4317 a random boundary is generated.
4319 Reference: https://tools.ietf.org/html/rfc7578
4321 has_specified_boundary
= boundary
is not None
4324 if boundary
is None:
4325 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
4328 out
, content_type
= _multipart_encode_impl(data
, boundary
)
4331 if has_specified_boundary
:
4335 return out
, content_type
4338 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
4339 if isinstance(key_or_keys
, (list, tuple)):
4340 for key
in key_or_keys
:
4341 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
4345 return d
.get(key_or_keys
, default
)
4348 def try_get(src
, getter
, expected_type
=None):
4349 for get
in variadic(getter
):
4352 except (AttributeError, KeyError, TypeError, IndexError):
4355 if expected_type
is None or isinstance(v
, expected_type
):
4359 def merge_dicts(*dicts
):
4361 for a_dict
in dicts
:
4362 for k
, v
in a_dict
.items():
4366 or (isinstance(v
, compat_str
) and v
4367 and isinstance(merged
[k
], compat_str
)
4368 and not merged
[k
])):
4373 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
4374 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
4386 TV_PARENTAL_GUIDELINES
= {
4396 def parse_age_limit(s
):
4398 return s
if 0 <= s
<= 21 else None
4399 if not isinstance(s
, compat_basestring
):
4401 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
4403 return int(m
.group('age'))
4406 return US_RATINGS
[s
]
4407 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
4409 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
4413 def strip_jsonp(code
):
4416 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
4417 (?:\s*&&\s*(?P=func_name))?
4418 \s*\(\s*(?P<callback_data>.*)\);?
4419 \s*?(?://[^\n]*)*$''',
4420 r
'\g<callback_data>', code
)
4423 def js_to_json(code
, vars={}):
4424 # vars is a dict of var, val pairs to substitute
4425 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
4426 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
4428 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
4429 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
4434 if v
in ('true', 'false', 'null'):
4436 elif v
in ('undefined', 'void 0'):
4438 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
4441 if v
[0] in ("'", '"'):
4442 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
4447 }.get(m
.group(0), m
.group(0)), v
[1:-1])
4449 for regex
, base
in INTEGER_TABLE
:
4450 im
= re
.match(regex
, v
)
4452 i
= int(im
.group(1), base
)
4453 return '"%d":' % i
if v
.endswith(':') else '%d' % i
4460 return re
.sub(r
'''(?sx)
4461 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
4462 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
4463 {comment}|,(?={skip}[\]}}])|
4464 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
4465 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
4468 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
4471 def qualities(quality_ids
):
4472 """ Get a numeric quality value out of a list of possible values """
4475 return quality_ids
.index(qid
)
4482 'default': '%(title)s [%(id)s].%(ext)s',
4483 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
4489 'description': 'description',
4490 'annotation': 'annotations.xml',
4491 'infojson': 'info.json',
4492 'pl_thumbnail': None,
4493 'pl_description': 'description',
4494 'pl_infojson': 'info.json',
4497 # As of [1] format syntax is:
4498 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
4499 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
4500 STR_FORMAT_RE_TMPL
= r
'''(?x)
4501 (?<!%)(?P<prefix>(?:%%)*)
4503 (?P<has_key>\((?P<key>{0})\))?
4505 (?P<conversion>[#0\-+ ]+)?
4507 (?P<precision>\.\d+)?
4508 (?P<len_mod>[hlL])? # unused in python
4509 {1} # conversion type
4514 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
4517 def limit_length(s
, length
):
4518 """ Add ellipses to overly long strings """
4523 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
4527 def version_tuple(v
):
4528 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
4531 def is_outdated_version(version
, limit
, assume_new
=True):
4533 return not assume_new
4535 return version_tuple(version
) < version_tuple(limit
)
4537 return not assume_new
4540 def ytdl_is_updateable():
4541 """ Returns if yt-dlp can be updated with -U """
4543 from .update
import is_non_updateable
4545 return not is_non_updateable()
4548 def args_to_str(args
):
4549 # Get a short string representation for a subprocess command
4550 return ' '.join(compat_shlex_quote(a
) for a
in args
)
4553 def error_to_compat_str(err
):
4555 # On python 2 error byte string must be decoded with proper
4556 # encoding rather than ascii
4557 if sys
.version_info
[0] < 3:
4558 err_str
= err_str
.decode(preferredencoding())
4562 def mimetype2ext(mt
):
4566 mt
, _
, params
= mt
.partition(';')
4571 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
4572 # it's the most popular one
4573 'audio/mpeg': 'mp3',
4574 'audio/x-wav': 'wav',
4576 'audio/wave': 'wav',
4579 ext
= FULL_MAP
.get(mt
)
4585 'smptett+xml': 'tt',
4589 'x-mp4-fragmented': 'mp4',
4590 'x-ms-sami': 'sami',
4593 'x-mpegurl': 'm3u8',
4594 'vnd.apple.mpegurl': 'm3u8',
4598 'vnd.ms-sstr+xml': 'ism',
4602 'filmstrip+json': 'fs',
4606 _
, _
, subtype
= mt
.rpartition('/')
4607 ext
= SUBTYPE_MAP
.get(subtype
.lower())
4618 _
, _
, suffix
= subtype
.partition('+')
4619 ext
= SUFFIX_MAP
.get(suffix
)
4623 return subtype
.replace('+', '.')
4626 def parse_codecs(codecs_str
):
4627 # http://tools.ietf.org/html/rfc6381
4630 split_codecs
= list(filter(None, map(
4631 str.strip
, codecs_str
.strip().strip(',').split(','))))
4632 vcodec
, acodec
, hdr
= None, None, None
4633 for full_codec
in split_codecs
:
4634 codec
= full_codec
.split('.')[0]
4635 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01', 'theora', 'dvh1', 'dvhe'):
4638 if codec
in ('dvh1', 'dvhe'):
4640 elif codec
== 'vp9' and vcodec
.startswith('vp9.2'):
4642 elif codec
== 'av01':
4643 parts
= full_codec
.split('.')
4644 if len(parts
) > 3 and parts
[3] == '10':
4646 vcodec
= '.'.join(parts
[:4])
4647 elif codec
in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
4651 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
4652 if not vcodec
and not acodec
:
4653 if len(split_codecs
) == 2:
4655 'vcodec': split_codecs
[0],
4656 'acodec': split_codecs
[1],
4660 'vcodec': vcodec
or 'none',
4661 'acodec': acodec
or 'none',
4662 'dynamic_range': hdr
,
4667 def urlhandle_detect_ext(url_handle
):
4668 getheader
= url_handle
.headers
.get
4670 cd
= getheader('Content-Disposition')
4672 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
4674 e
= determine_ext(m
.group('filename'), default_ext
=None)
4678 return mimetype2ext(getheader('Content-Type'))
4681 def encode_data_uri(data
, mime_type
):
4682 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
4685 def age_restricted(content_limit
, age_limit
):
4686 """ Returns True iff the content should be blocked """
4688 if age_limit
is None: # No limit set
4690 if content_limit
is None:
4691 return False # Content available for everyone
4692 return age_limit
< content_limit
4695 def is_html(first_bytes
):
4696 """ Detect whether a file contains HTML by examining its first bytes. """
4699 (b
'\xef\xbb\xbf', 'utf-8'),
4700 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
4701 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
4702 (b
'\xff\xfe', 'utf-16-le'),
4703 (b
'\xfe\xff', 'utf-16-be'),
4705 for bom
, enc
in BOMS
:
4706 if first_bytes
.startswith(bom
):
4707 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
4710 s
= first_bytes
.decode('utf-8', 'replace')
4712 return re
.match(r
'^\s*<', s
)
4715 def determine_protocol(info_dict
):
4716 protocol
= info_dict
.get('protocol')
4717 if protocol
is not None:
4720 url
= info_dict
['url']
4721 if url
.startswith('rtmp'):
4723 elif url
.startswith('mms'):
4725 elif url
.startswith('rtsp'):
4728 ext
= determine_ext(url
)
4734 return compat_urllib_parse_urlparse(url
).scheme
4737 def render_table(header_row
, data
, delim
=False, extraGap
=0, hideEmpty
=False):
4738 """ Render a list of rows, each as a list of values """
4740 def get_max_lens(table
):
4741 return [max(len(compat_str(v
)) for v
in col
) for col
in zip(*table
)]
4743 def filter_using_list(row
, filterArray
):
4744 return [col
for (take
, col
) in zip(filterArray
, row
) if take
]
4747 max_lens
= get_max_lens(data
)
4748 header_row
= filter_using_list(header_row
, max_lens
)
4749 data
= [filter_using_list(row
, max_lens
) for row
in data
]
4751 table
= [header_row
] + data
4752 max_lens
= get_max_lens(table
)
4754 table
= [header_row
] + [['-' * ml
for ml
in max_lens
]] + data
4755 format_str
= ' '.join('%-' + compat_str(ml
+ extraGap
) + 's' for ml
in max_lens
[:-1]) + ' %s'
4756 return '\n'.join(format_str
% tuple(row
) for row
in table
)
4759 def _match_one(filter_part
, dct
, incomplete
):
4760 # TODO: Generalize code with YoutubeDL._build_format_filter
4761 STRING_OPERATORS
= {
4762 '*=': operator
.contains
,
4763 '^=': lambda attr
, value
: attr
.startswith(value
),
4764 '$=': lambda attr
, value
: attr
.endswith(value
),
4765 '~=': lambda attr
, value
: re
.search(value
, attr
),
4767 COMPARISON_OPERATORS
= {
4769 '<=': operator
.le
, # "<=" must be defined above "<"
4776 operator_rex
= re
.compile(r
'''(?x)\s*
4778 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
4780 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
4784 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
4785 m = operator_rex.search(filter_part)
4788 unnegated_op = COMPARISON_OPERATORS[m['op']]
4790 op = lambda attr, value: not unnegated_op(attr, value)
4793 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
4795 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
4796 actual_value = dct.get(m['key'])
4797 numeric_comparison = None
4798 if isinstance(actual_value, compat_numeric_types):
4799 # If the original field is a string and matching comparisonvalue is
4800 # a number we should respect the origin of the original field
4801 # and process comparison value as a string (see
4802 # https://github.com/ytdl-org/youtube-dl/issues/11082)
4804 numeric_comparison = int(comparison_value)
4806 numeric_comparison = parse_filesize(comparison_value)
4807 if numeric_comparison is None:
4808 numeric_comparison = parse_filesize(f'{comparison_value}B')
4809 if numeric_comparison is None:
4810 numeric_comparison = parse_duration(comparison_value)
4811 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
4812 raise ValueError('Operator %s only supports string values!' % m['op'])
4813 if actual_value is None:
4814 return incomplete or m['none_inclusive']
4815 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
4818 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
4819 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
4821 operator_rex = re.compile(r'''(?x
)\s
*
4822 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
4824 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
4825 m = operator_rex.search(filter_part)
4827 op = UNARY_OPERATORS[m.group('op')]
4828 actual_value = dct.get(m.group('key'))
4829 if incomplete and actual_value is None:
4831 return op(actual_value)
4833 raise ValueError('Invalid filter part %r' % filter_part)
4836 def match_str(filter_str, dct, incomplete=False):
4837 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
4838 When incomplete, all conditions passes on missing fields
4841 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
4842 for filter_part in re.split(r'(?<!\\)&', filter_str))
4845 def match_filter_func(filter_str):
4846 def _match_func(info_dict, *args, **kwargs):
4847 if match_str(filter_str, info_dict, *args, **kwargs):
4850 video_title = info_dict.get('title', info_dict.get('id', 'video'))
4851 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
4855 def parse_dfxp_time_expr(time_expr):
4859 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
4861 return float(mobj.group('time_offset'))
4863 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
4865 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
4868 def srt_subtitles_timecode(seconds):
4869 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
4872 def ass_subtitles_timecode(seconds):
4873 time = timetuple_from_msec(seconds * 1000)
4874 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
4877 def dfxp2srt(dfxp_data):
4879 @param dfxp_data A
bytes-like
object containing DFXP data
4880 @returns A
unicode object containing converted SRT data
4882 LEGACY_NAMESPACES = (
4883 (b'http://www.w3.org/ns/ttml', [
4884 b'http://www.w3.org/2004/11/ttaf1',
4885 b'http://www.w3.org/2006/04/ttaf1',
4886 b'http://www.w3.org/2006/10/ttaf1',
4888 (b'http://www.w3.org/ns/ttml#styling', [
4889 b'http://www.w3.org/ns/ttml#style',
4893 SUPPORTED_STYLING = [
4902 _x = functools.partial(xpath_with_ns, ns_map={
4903 'xml': 'http://www.w3.org/XML/1998/namespace',
4904 'ttml': 'http://www.w3.org/ns/ttml',
4905 'tts': 'http://www.w3.org/ns/ttml#styling',
4911 class TTMLPElementParser(object):
4913 _unclosed_elements = []
4914 _applied_styles = []
4916 def start(self, tag, attrib):
4917 if tag in (_x('ttml:br'), 'br'):
4920 unclosed_elements = []
4922 element_style_id = attrib.get('style')
4924 style.update(default_style)
4925 if element_style_id:
4926 style.update(styles.get(element_style_id, {}))
4927 for prop in SUPPORTED_STYLING:
4928 prop_val = attrib.get(_x('tts:' + prop))
4930 style[prop] = prop_val
4933 for k, v in sorted(style.items()):
4934 if self._applied_styles and self._applied_styles[-1].get(k) == v:
4937 font += ' color="%s"' % v
4938 elif k == 'fontSize':
4939 font += ' size="%s"' % v
4940 elif k == 'fontFamily':
4941 font += ' face="%s"' % v
4942 elif k == 'fontWeight' and v == 'bold':
4944 unclosed_elements.append('b')
4945 elif k == 'fontStyle' and v == 'italic':
4947 unclosed_elements.append('i')
4948 elif k == 'textDecoration' and v == 'underline':
4950 unclosed_elements.append('u')
4952 self._out += '<font' + font + '>'
4953 unclosed_elements.append('font')
4955 if self._applied_styles:
4956 applied_style.update(self._applied_styles[-1])
4957 applied_style.update(style)
4958 self._applied_styles.append(applied_style)
4959 self._unclosed_elements.append(unclosed_elements)
4962 if tag not in (_x('ttml:br'), 'br'):
4963 unclosed_elements = self._unclosed_elements.pop()
4964 for element in reversed(unclosed_elements):
4965 self._out += '</%s>' % element
4966 if unclosed_elements and self._applied_styles:
4967 self._applied_styles.pop()
4969 def data(self, data):
4973 return self._out.strip()
4975 def parse_node(node):
4976 target = TTMLPElementParser()
4977 parser = xml.etree.ElementTree.XMLParser(target=target)
4978 parser.feed(xml.etree.ElementTree.tostring(node))
4979 return parser.close()
4981 for k, v in LEGACY_NAMESPACES:
4983 dfxp_data = dfxp_data.replace(ns, k)
4985 dfxp = compat_etree_fromstring(dfxp_data)
4987 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
4990 raise ValueError('Invalid dfxp/TTML subtitle')
4994 for style in dfxp.findall(_x('.//ttml:style')):
4995 style_id = style.get('id') or style.get(_x('xml:id'))
4998 parent_style_id = style.get('style')
5000 if parent_style_id not in styles:
5003 styles[style_id] = styles[parent_style_id].copy()
5004 for prop in SUPPORTED_STYLING:
5005 prop_val = style.get(_x('tts:' + prop))
5007 styles.setdefault(style_id, {})[prop] = prop_val
5013 for p in ('body', 'div'):
5014 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
5017 style = styles.get(ele.get('style'))
5020 default_style.update(style)
5022 for para, index in zip(paras, itertools.count(1)):
5023 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
5024 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
5025 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
5026 if begin_time is None:
5031 end_time = begin_time + dur
5032 out.append('%d\n%s --> %s\n%s\n\n' % (
5034 srt_subtitles_timecode(begin_time),
5035 srt_subtitles_timecode(end_time),
5041 def cli_option(params, command_option, param):
5042 param = params.get(param)
5044 param = compat_str(param)
5045 return [command_option, param] if param is not None else []
5048 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
5049 param = params.get(param)
5052 assert isinstance(param, bool)
5054 return [command_option + separator + (true_value if param else false_value)]
5055 return [command_option, true_value if param else false_value]
5058 def cli_valueless_option(params, command_option, param, expected_value=True):
5059 param = params.get(param)
5060 return [command_option] if param == expected_value else []
5063 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
5064 if isinstance(argdict, (list, tuple)): # for backward compatibility
5071 assert isinstance(argdict, dict)
5073 assert isinstance(keys, (list, tuple))
5074 for key_list in keys:
5075 arg_list = list(filter(
5076 lambda x: x is not None,
5077 [argdict.get(key.lower()) for key in variadic(key_list)]))
5079 return [arg for args in arg_list for arg in args]
5083 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
5084 main_key, exe = main_key.lower(), exe.lower()
5085 root_key = exe if main_key == exe else f'{main_key}+{exe}'
5086 keys = [f'{root_key}{k}' for k in (keys or [''])]
5087 if root_key in keys:
5089 keys.append((main_key, exe))
5090 keys.append('default')
5093 return cli_configuration_args(argdict, keys, default, use_compat)
5096 class ISO639Utils(object):
5097 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
5156 'iw': 'heb', # Replaced by he in 1989 revision
5166 'in': 'ind', # Replaced by id in 1989 revision
5281 'ji': 'yid', # Replaced by yi in 1989 revision
5289 def short2long(cls, code):
5290 """Convert language code from ISO 639-1 to ISO 639-2/T"""
5291 return cls._lang_map.get(code[:2])
5294 def long2short(cls, code):
5295 """Convert language code from ISO 639-2/T to ISO 639-1"""
5296 for short_name, long_name in cls._lang_map.items():
5297 if long_name == code:
5301 class ISO3166Utils(object):
5302 # From http://data.okfn.org/data/core/country-list
5304 'AF': 'Afghanistan',
5305 'AX': 'Åland Islands',
5308 'AS': 'American Samoa',
5313 'AG': 'Antigua and Barbuda',
5330 'BO': 'Bolivia, Plurinational State of',
5331 'BQ': 'Bonaire, Sint Eustatius and Saba',
5332 'BA': 'Bosnia and Herzegovina',
5334 'BV': 'Bouvet Island',
5336 'IO': 'British Indian Ocean Territory',
5337 'BN': 'Brunei Darussalam',
5339 'BF': 'Burkina Faso',
5345 'KY': 'Cayman Islands',
5346 'CF': 'Central African Republic',
5350 'CX': 'Christmas Island',
5351 'CC': 'Cocos (Keeling) Islands',
5355 'CD': 'Congo, the Democratic Republic of the',
5356 'CK': 'Cook Islands',
5358 'CI': 'Côte d\'Ivoire',
5363 'CZ': 'Czech Republic',
5367 'DO': 'Dominican Republic',
5370 'SV': 'El Salvador',
5371 'GQ': 'Equatorial Guinea',
5375 'FK': 'Falkland Islands (Malvinas)',
5376 'FO': 'Faroe Islands',
5380 'GF': 'French Guiana',
5381 'PF': 'French Polynesia',
5382 'TF': 'French Southern Territories',
5397 'GW': 'Guinea-Bissau',
5400 'HM': 'Heard Island and McDonald Islands',
5401 'VA': 'Holy See (Vatican City State)',
5408 'IR': 'Iran, Islamic Republic of',
5411 'IM': 'Isle of Man',
5421 'KP': 'Korea, Democratic People\'s Republic of',
5422 'KR': 'Korea, Republic of',
5425 'LA': 'Lao People\'s Democratic Republic',
5431 'LI': 'Liechtenstein',
5435 'MK': 'Macedonia, the Former Yugoslav Republic of',
5442 'MH': 'Marshall Islands',
5448 'FM': 'Micronesia, Federated States of',
5449 'MD': 'Moldova, Republic of',
5460 'NL': 'Netherlands',
5461 'NC': 'New Caledonia',
5462 'NZ': 'New Zealand',
5467 'NF': 'Norfolk Island',
5468 'MP': 'Northern Mariana Islands',
5473 'PS': 'Palestine, State of',
5475 'PG': 'Papua New Guinea',
5478 'PH': 'Philippines',
5482 'PR': 'Puerto Rico',
5486 'RU': 'Russian Federation',
5488 'BL': 'Saint Barthélemy',
5489 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
5490 'KN': 'Saint Kitts and Nevis',
5491 'LC': 'Saint Lucia',
5492 'MF': 'Saint Martin (French part)',
5493 'PM': 'Saint Pierre and Miquelon',
5494 'VC': 'Saint Vincent and the Grenadines',
5497 'ST': 'Sao Tome and Principe',
5498 'SA': 'Saudi Arabia',
5502 'SL': 'Sierra Leone',
5504 'SX': 'Sint Maarten (Dutch part)',
5507 'SB': 'Solomon Islands',
5509 'ZA': 'South Africa',
5510 'GS': 'South Georgia and the South Sandwich Islands',
5511 'SS': 'South Sudan',
5516 'SJ': 'Svalbard and Jan Mayen',
5519 'CH': 'Switzerland',
5520 'SY': 'Syrian Arab Republic',
5521 'TW': 'Taiwan, Province of China',
5523 'TZ': 'Tanzania, United Republic of',
5525 'TL': 'Timor-Leste',
5529 'TT': 'Trinidad and Tobago',
5532 'TM': 'Turkmenistan',
5533 'TC': 'Turks and Caicos Islands',
5537 'AE': 'United Arab Emirates',
5538 'GB': 'United Kingdom',
5539 'US': 'United States',
5540 'UM': 'United States Minor Outlying Islands',
5544 'VE': 'Venezuela, Bolivarian Republic of',
5546 'VG': 'Virgin Islands, British',
5547 'VI': 'Virgin Islands, U.S.',
5548 'WF': 'Wallis and Futuna',
5549 'EH': 'Western Sahara',
5556 def short2full(cls, code):
5557 """Convert an ISO 3166-2 country code to the corresponding full name"""
5558 return cls._country_map.get(code.upper())
5561 class GeoUtils(object):
5562 # Major IPv4 address blocks per country
5564 'AD': '46.172.224.0/19',
5565 'AE': '94.200.0.0/13',
5566 'AF': '149.54.0.0/17',
5567 'AG': '209.59.64.0/18',
5568 'AI': '204.14.248.0/21',
5569 'AL': '46.99.0.0/16',
5570 'AM': '46.70.0.0/15',
5571 'AO': '105.168.0.0/13',
5572 'AP': '182.50.184.0/21',
5573 'AQ': '23.154.160.0/24',
5574 'AR': '181.0.0.0/12',
5575 'AS': '202.70.112.0/20',
5576 'AT': '77.116.0.0/14',
5577 'AU': '1.128.0.0/11',
5578 'AW': '181.41.0.0/18',
5579 'AX': '185.217.4.0/22',
5580 'AZ': '5.197.0.0/16',
5581 'BA': '31.176.128.0/17',
5582 'BB': '65.48.128.0/17',
5583 'BD': '114.130.0.0/16',
5585 'BF': '102.178.0.0/15',
5586 'BG': '95.42.0.0/15',
5587 'BH': '37.131.0.0/17',
5588 'BI': '154.117.192.0/18',
5589 'BJ': '137.255.0.0/16',
5590 'BL': '185.212.72.0/23',
5591 'BM': '196.12.64.0/18',
5592 'BN': '156.31.0.0/16',
5593 'BO': '161.56.0.0/16',
5594 'BQ': '161.0.80.0/20',
5595 'BR': '191.128.0.0/12',
5596 'BS': '24.51.64.0/18',
5597 'BT': '119.2.96.0/19',
5598 'BW': '168.167.0.0/16',
5599 'BY': '178.120.0.0/13',
5600 'BZ': '179.42.192.0/18',
5601 'CA': '99.224.0.0/11',
5602 'CD': '41.243.0.0/16',
5603 'CF': '197.242.176.0/21',
5604 'CG': '160.113.0.0/16',
5605 'CH': '85.0.0.0/13',
5606 'CI': '102.136.0.0/14',
5607 'CK': '202.65.32.0/19',
5608 'CL': '152.172.0.0/14',
5609 'CM': '102.244.0.0/14',
5610 'CN': '36.128.0.0/10',
5611 'CO': '181.240.0.0/12',
5612 'CR': '201.192.0.0/12',
5613 'CU': '152.206.0.0/15',
5614 'CV': '165.90.96.0/19',
5615 'CW': '190.88.128.0/17',
5616 'CY': '31.153.0.0/16',
5617 'CZ': '88.100.0.0/14',
5619 'DJ': '197.241.0.0/17',
5620 'DK': '87.48.0.0/12',
5621 'DM': '192.243.48.0/20',
5622 'DO': '152.166.0.0/15',
5623 'DZ': '41.96.0.0/12',
5624 'EC': '186.68.0.0/15',
5625 'EE': '90.190.0.0/15',
5626 'EG': '156.160.0.0/11',
5627 'ER': '196.200.96.0/20',
5628 'ES': '88.0.0.0/11',
5629 'ET': '196.188.0.0/14',
5630 'EU': '2.16.0.0/13',
5631 'FI': '91.152.0.0/13',
5632 'FJ': '144.120.0.0/16',
5633 'FK': '80.73.208.0/21',
5634 'FM': '119.252.112.0/20',
5635 'FO': '88.85.32.0/19',
5637 'GA': '41.158.0.0/15',
5639 'GD': '74.122.88.0/21',
5640 'GE': '31.146.0.0/16',
5641 'GF': '161.22.64.0/18',
5642 'GG': '62.68.160.0/19',
5643 'GH': '154.160.0.0/12',
5644 'GI': '95.164.0.0/16',
5645 'GL': '88.83.0.0/19',
5646 'GM': '160.182.0.0/15',
5647 'GN': '197.149.192.0/18',
5648 'GP': '104.250.0.0/19',
5649 'GQ': '105.235.224.0/20',
5650 'GR': '94.64.0.0/13',
5651 'GT': '168.234.0.0/16',
5652 'GU': '168.123.0.0/16',
5653 'GW': '197.214.80.0/20',
5654 'GY': '181.41.64.0/18',
5655 'HK': '113.252.0.0/14',
5656 'HN': '181.210.0.0/16',
5657 'HR': '93.136.0.0/13',
5658 'HT': '148.102.128.0/17',
5659 'HU': '84.0.0.0/14',
5660 'ID': '39.192.0.0/10',
5661 'IE': '87.32.0.0/12',
5662 'IL': '79.176.0.0/13',
5663 'IM': '5.62.80.0/20',
5664 'IN': '117.192.0.0/10',
5665 'IO': '203.83.48.0/21',
5666 'IQ': '37.236.0.0/14',
5667 'IR': '2.176.0.0/12',
5668 'IS': '82.221.0.0/16',
5669 'IT': '79.0.0.0/10',
5670 'JE': '87.244.64.0/18',
5671 'JM': '72.27.0.0/17',
5672 'JO': '176.29.0.0/16',
5673 'JP': '133.0.0.0/8',
5674 'KE': '105.48.0.0/12',
5675 'KG': '158.181.128.0/17',
5676 'KH': '36.37.128.0/17',
5677 'KI': '103.25.140.0/22',
5678 'KM': '197.255.224.0/20',
5679 'KN': '198.167.192.0/19',
5680 'KP': '175.45.176.0/22',
5681 'KR': '175.192.0.0/10',
5682 'KW': '37.36.0.0/14',
5683 'KY': '64.96.0.0/15',
5684 'KZ': '2.72.0.0/13',
5685 'LA': '115.84.64.0/18',
5686 'LB': '178.135.0.0/16',
5687 'LC': '24.92.144.0/20',
5688 'LI': '82.117.0.0/19',
5689 'LK': '112.134.0.0/15',
5690 'LR': '102.183.0.0/16',
5691 'LS': '129.232.0.0/17',
5692 'LT': '78.56.0.0/13',
5693 'LU': '188.42.0.0/16',
5694 'LV': '46.109.0.0/16',
5695 'LY': '41.252.0.0/14',
5696 'MA': '105.128.0.0/11',
5697 'MC': '88.209.64.0/18',
5698 'MD': '37.246.0.0/16',
5699 'ME': '178.175.0.0/17',
5700 'MF': '74.112.232.0/21',
5701 'MG': '154.126.0.0/17',
5702 'MH': '117.103.88.0/21',
5703 'MK': '77.28.0.0/15',
5704 'ML': '154.118.128.0/18',
5705 'MM': '37.111.0.0/17',
5706 'MN': '49.0.128.0/17',
5707 'MO': '60.246.0.0/16',
5708 'MP': '202.88.64.0/20',
5709 'MQ': '109.203.224.0/19',
5710 'MR': '41.188.64.0/18',
5711 'MS': '208.90.112.0/22',
5712 'MT': '46.11.0.0/16',
5713 'MU': '105.16.0.0/12',
5714 'MV': '27.114.128.0/18',
5715 'MW': '102.70.0.0/15',
5716 'MX': '187.192.0.0/11',
5717 'MY': '175.136.0.0/13',
5718 'MZ': '197.218.0.0/15',
5719 'NA': '41.182.0.0/16',
5720 'NC': '101.101.0.0/18',
5721 'NE': '197.214.0.0/18',
5722 'NF': '203.17.240.0/22',
5723 'NG': '105.112.0.0/12',
5724 'NI': '186.76.0.0/15',
5725 'NL': '145.96.0.0/11',
5726 'NO': '84.208.0.0/13',
5727 'NP': '36.252.0.0/15',
5728 'NR': '203.98.224.0/19',
5729 'NU': '49.156.48.0/22',
5730 'NZ': '49.224.0.0/14',
5731 'OM': '5.36.0.0/15',
5732 'PA': '186.72.0.0/15',
5733 'PE': '186.160.0.0/14',
5734 'PF': '123.50.64.0/18',
5735 'PG': '124.240.192.0/19',
5736 'PH': '49.144.0.0/13',
5737 'PK': '39.32.0.0/11',
5738 'PL': '83.0.0.0/11',
5739 'PM': '70.36.0.0/20',
5740 'PR': '66.50.0.0/16',
5741 'PS': '188.161.0.0/16',
5742 'PT': '85.240.0.0/13',
5743 'PW': '202.124.224.0/20',
5744 'PY': '181.120.0.0/14',
5745 'QA': '37.210.0.0/15',
5746 'RE': '102.35.0.0/16',
5747 'RO': '79.112.0.0/13',
5748 'RS': '93.86.0.0/15',
5749 'RU': '5.136.0.0/13',
5750 'RW': '41.186.0.0/16',
5751 'SA': '188.48.0.0/13',
5752 'SB': '202.1.160.0/19',
5753 'SC': '154.192.0.0/11',
5754 'SD': '102.120.0.0/13',
5755 'SE': '78.64.0.0/12',
5756 'SG': '8.128.0.0/10',
5757 'SI': '188.196.0.0/14',
5758 'SK': '78.98.0.0/15',
5759 'SL': '102.143.0.0/17',
5760 'SM': '89.186.32.0/19',
5761 'SN': '41.82.0.0/15',
5762 'SO': '154.115.192.0/18',
5763 'SR': '186.179.128.0/17',
5764 'SS': '105.235.208.0/21',
5765 'ST': '197.159.160.0/19',
5766 'SV': '168.243.0.0/16',
5767 'SX': '190.102.0.0/20',
5769 'SZ': '41.84.224.0/19',
5770 'TC': '65.255.48.0/20',
5771 'TD': '154.68.128.0/19',
5772 'TG': '196.168.0.0/14',
5773 'TH': '171.96.0.0/13',
5774 'TJ': '85.9.128.0/18',
5775 'TK': '27.96.24.0/21',
5776 'TL': '180.189.160.0/20',
5777 'TM': '95.85.96.0/19',
5778 'TN': '197.0.0.0/11',
5779 'TO': '175.176.144.0/21',
5780 'TR': '78.160.0.0/11',
5781 'TT': '186.44.0.0/15',
5782 'TV': '202.2.96.0/19',
5783 'TW': '120.96.0.0/11',
5784 'TZ': '156.156.0.0/14',
5785 'UA': '37.52.0.0/14',
5786 'UG': '102.80.0.0/13',
5788 'UY': '167.56.0.0/13',
5789 'UZ': '84.54.64.0/18',
5790 'VA': '212.77.0.0/19',
5791 'VC': '207.191.240.0/21',
5792 'VE': '186.88.0.0/13',
5793 'VG': '66.81.192.0/20',
5794 'VI': '146.226.0.0/16',
5795 'VN': '14.160.0.0/11',
5796 'VU': '202.80.32.0/20',
5797 'WF': '117.20.32.0/21',
5798 'WS': '202.4.32.0/19',
5799 'YE': '134.35.0.0/16',
5800 'YT': '41.242.116.0/22',
5801 'ZA': '41.0.0.0/11',
5802 'ZM': '102.144.0.0/13',
5803 'ZW': '102.177.192.0/18',
5807 def random_ipv4(cls, code_or_block):
5808 if len(code_or_block) == 2:
5809 block = cls._country_ip_map.get(code_or_block.upper())
5813 block = code_or_block
5814 addr, preflen = block.split('/')
5815 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
5816 addr_max = addr_min | (0xffffffff >> int(preflen))
5817 return compat_str(socket.inet_ntoa(
5818 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
5821 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
5822 def __init__(self, proxies=None):
5823 # Set default handlers
5824 for type in ('http', 'https'):
5825 setattr(self, '%s_open' % type,
5826 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
5827 meth(r, proxy, type))
5828 compat_urllib_request.ProxyHandler.__init__(self, proxies)
5830 def proxy_open(self, req, proxy, type):
5831 req_proxy = req.headers.get('Ytdl-request-proxy')
5832 if req_proxy is not None:
5834 del req.headers['Ytdl-request-proxy']
5836 if proxy == '__noproxy__':
5837 return None # No Proxy
5838 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
5839 req.add_header('Ytdl-socks-proxy', proxy)
5840 # yt-dlp's http/https handlers do wrapping the socket with socks
5842 return compat_urllib_request.ProxyHandler.proxy_open(
5843 self, req, proxy, type)
5846 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
5847 # released into Public Domain
5848 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
5850 def long_to_bytes(n, blocksize=0):
5851 """long_to_bytes(n:long, blocksize:int) : string
5852 Convert a long integer to a byte string.
5854 If optional blocksize is given and greater than zero, pad the front of the
5855 byte string with binary zeros so that the length is a multiple of
5858 # after much testing, this algorithm was deemed to be the fastest
5862 s = compat_struct_pack('>I', n & 0xffffffff) + s
5864 # strip off leading zeros
5865 for i in range(len(s)):
5866 if s[i] != b'\000'[0]:
5869 # only happens when n == 0
5873 # add back some pad bytes. this could be done more efficiently w.r.t. the
5874 # de-padding being done above, but sigh...
5875 if blocksize > 0 and len(s) % blocksize:
5876 s = (blocksize - len(s) % blocksize) * b'\000' + s
5880 def bytes_to_long(s):
5881 """bytes_to_long(string) : long
5882 Convert a byte string to a long integer.
5884 This is (essentially) the inverse of long_to_bytes().
5889 extra = (4 - length % 4)
5890 s = b'\000' * extra + s
5891 length = length + extra
5892 for i in range(0, length, 4):
5893 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
5897 def ohdave_rsa_encrypt(data, exponent, modulus):
5899 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
5902 data: data to encrypt, bytes-like object
5903 exponent, modulus: parameter e and N of RSA algorithm, both integer
5904 Output: hex string of encrypted data
5906 Limitation: supports one block encryption only
5909 payload = int(binascii.hexlify(data[::-1]), 16)
5910 encrypted = pow(payload, exponent, modulus)
5911 return '%x' % encrypted
5914 def pkcs1pad(data, length):
5916 Padding input data with PKCS#1 scheme
5918 @param {int[]} data input data
5919 @param {int} length target length
5920 @returns {int[]} padded data
5922 if len(data) > length - 11:
5923 raise ValueError('Input data too
long for PKCS
#1 padding')
5925 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
5926 return [0, 2] + pseudo_random
+ [0] + data
5929 def encode_base_n(num
, n
, table
=None):
5930 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
5932 table
= FULL_TABLE
[:n
]
5935 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
5942 ret
= table
[num
% n
] + ret
5947 def decode_packed_codes(code
):
5948 mobj
= re
.search(PACKED_CODES_RE
, code
)
5949 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
5952 symbols
= symbols
.split('|')
5957 base_n_count
= encode_base_n(count
, base
)
5958 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
5961 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
5965 def caesar(s
, alphabet
, shift
):
5970 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
5975 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
5978 def parse_m3u8_attributes(attrib
):
5980 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
5981 if val
.startswith('"'):
5987 def urshift(val
, n
):
5988 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
5991 # Based on png2str() written by @gdkchan and improved by @yokrysty
5992 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
5993 def decode_png(png_data
):
5994 # Reference: https://www.w3.org/TR/PNG/
5995 header
= png_data
[8:]
5997 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
5998 raise IOError('Not a valid PNG file.')
6000 int_map
= {1: '>B', 2: '>H', 4: '>I'}
6001 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
6006 length
= unpack_integer(header
[:4])
6009 chunk_type
= header
[:4]
6012 chunk_data
= header
[:length
]
6013 header
= header
[length
:]
6015 header
= header
[4:] # Skip CRC
6023 ihdr
= chunks
[0]['data']
6025 width
= unpack_integer(ihdr
[:4])
6026 height
= unpack_integer(ihdr
[4:8])
6030 for chunk
in chunks
:
6031 if chunk
['type'] == b
'IDAT':
6032 idat
+= chunk
['data']
6035 raise IOError('Unable to read PNG data.')
6037 decompressed_data
= bytearray(zlib
.decompress(idat
))
6042 def _get_pixel(idx
):
6047 for y
in range(height
):
6048 basePos
= y
* (1 + stride
)
6049 filter_type
= decompressed_data
[basePos
]
6053 pixels
.append(current_row
)
6055 for x
in range(stride
):
6056 color
= decompressed_data
[1 + basePos
+ x
]
6057 basex
= y
* stride
+ x
6062 left
= _get_pixel(basex
- 3)
6064 up
= _get_pixel(basex
- stride
)
6066 if filter_type
== 1: # Sub
6067 color
= (color
+ left
) & 0xff
6068 elif filter_type
== 2: # Up
6069 color
= (color
+ up
) & 0xff
6070 elif filter_type
== 3: # Average
6071 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
6072 elif filter_type
== 4: # Paeth
6078 c
= _get_pixel(basex
- stride
- 3)
6086 if pa
<= pb
and pa
<= pc
:
6087 color
= (color
+ a
) & 0xff
6089 color
= (color
+ b
) & 0xff
6091 color
= (color
+ c
) & 0xff
6093 current_row
.append(color
)
6095 return width
, height
, pixels
6098 def write_xattr(path
, key
, value
):
6099 # This mess below finds the best xattr tool for the job
6101 # try the pyxattr module...
6104 if hasattr(xattr
, 'set'): # pyxattr
6105 # Unicode arguments are not supported in python-pyxattr until
6107 # See https://github.com/ytdl-org/youtube-dl/issues/5498
6108 pyxattr_required_version
= '0.5.0'
6109 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
6110 # TODO: fallback to CLI tools
6111 raise XAttrUnavailableError(
6112 'python-pyxattr is detected but is too old. '
6113 'yt-dlp requires %s or above while your version is %s. '
6114 'Falling back to other xattr implementations' % (
6115 pyxattr_required_version
, xattr
.__version
__))
6117 setxattr
= xattr
.set
6119 setxattr
= xattr
.setxattr
6122 setxattr(path
, key
, value
)
6123 except EnvironmentError as e
:
6124 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6127 if compat_os_name
== 'nt':
6128 # Write xattrs to NTFS Alternate Data Streams:
6129 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
6130 assert ':' not in key
6131 assert os
.path
.exists(path
)
6133 ads_fn
= path
+ ':' + key
6135 with open(ads_fn
, 'wb') as f
:
6137 except EnvironmentError as e
:
6138 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6140 user_has_setfattr
= check_executable('setfattr', ['--version'])
6141 user_has_xattr
= check_executable('xattr', ['-h'])
6143 if user_has_setfattr
or user_has_xattr
:
6145 value
= value
.decode('utf-8')
6146 if user_has_setfattr
:
6147 executable
= 'setfattr'
6148 opts
= ['-n', key
, '-v', value
]
6149 elif user_has_xattr
:
6150 executable
= 'xattr'
6151 opts
= ['-w', key
, value
]
6153 cmd
= ([encodeFilename(executable
, True)]
6154 + [encodeArgument(o
) for o
in opts
]
6155 + [encodeFilename(path
, True)])
6158 p
= subprocess
.Popen(
6159 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
6160 except EnvironmentError as e
:
6161 raise XAttrMetadataError(e
.errno
, e
.strerror
)
6162 stdout
, stderr
= process_communicate_or_kill(p
)
6163 stderr
= stderr
.decode('utf-8', 'replace')
6164 if p
.returncode
!= 0:
6165 raise XAttrMetadataError(p
.returncode
, stderr
)
6168 # On Unix, and can't find pyxattr, setfattr, or xattr.
6169 if sys
.platform
.startswith('linux'):
6170 raise XAttrUnavailableError(
6171 "Couldn't find a tool to set the xattrs. "
6172 "Install either the python 'pyxattr' or 'xattr' "
6173 "modules, or the GNU 'attr' package "
6174 "(which contains the 'setfattr' tool).")
6176 raise XAttrUnavailableError(
6177 "Couldn't find a tool to set the xattrs. "
6178 "Install either the python 'xattr' module, "
6179 "or the 'xattr' binary.")
6182 def random_birthday(year_field
, month_field
, day_field
):
6183 start_date
= datetime
.date(1950, 1, 1)
6184 end_date
= datetime
.date(1995, 12, 31)
6185 offset
= random
.randint(0, (end_date
- start_date
).days
)
6186 random_date
= start_date
+ datetime
.timedelta(offset
)
6188 year_field
: str(random_date
.year
),
6189 month_field
: str(random_date
.month
),
6190 day_field
: str(random_date
.day
),
6194 # Templates for internet shortcut files, which are plain text files.
6195 DOT_URL_LINK_TEMPLATE
= '''
6200 DOT_WEBLOC_LINK_TEMPLATE
= '''
6201 <?xml version="1.0" encoding="UTF-8"?>
6202 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
6203 <plist version="1.0">
6206 \t<string>%(url)s</string>
6211 DOT_DESKTOP_LINK_TEMPLATE
= '''
6221 def iri_to_uri(iri
):
6223 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
6225 The function doesn't add an additional layer of escaping; e.g., it doesn't escape `%3C` as `%253C`. Instead, it percent-escapes characters with an underlying UTF-8 encoding *besides* those already escaped, leaving the URI intact.
6228 iri_parts
= compat_urllib_parse_urlparse(iri
)
6230 if '[' in iri_parts
.netloc
:
6231 raise ValueError('IPv6 URIs are not, yet, supported.')
6232 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
6234 # The `safe` argument values, that the following code uses, contain the characters that should not be percent-encoded. Everything else but letters, digits and '_.-' will be percent-encoded with an underlying UTF-8 encoding. Everything already percent-encoded will be left as is.
6237 if iri_parts
.username
:
6238 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
6239 if iri_parts
.password
is not None:
6240 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
6243 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
6244 # The 'idna' encoding produces ASCII text.
6245 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
6246 net_location
+= ':' + str(iri_parts
.port
)
6248 return compat_urllib_parse_urlunparse(
6252 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
6254 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
6255 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
6257 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
6258 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
6260 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
6262 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
6265 def to_high_limit_path(path
):
6266 if sys
.platform
in ['win32', 'cygwin']:
6267 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
6268 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
6273 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
6275 val
= obj
if obj
is not None else default
6277 val
= obj
.get(field
, default
)
6278 if func
and val
not in ignore
:
6280 return template
% val
if val
not in ignore
else default
6283 def clean_podcast_url(url
):
6284 return re
.sub(r
'''(?x)
6288 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
6291 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
6294 cn\.co| # https://podcorn.com/analytics-prefix/
6295 st\.fm # https://podsights.com/docs/
6300 _HEX_TABLE
= '0123456789abcdef'
6303 def random_uuidv4():
6304 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
6307 def make_dir(path
, to_screen
=None):
6309 dn
= os
.path
.dirname(path
)
6310 if dn
and not os
.path
.exists(dn
):
6313 except (OSError, IOError) as err
:
6314 if callable(to_screen
) is not None:
6315 to_screen('unable to create directory ' + error_to_compat_str(err
))
6319 def get_executable_path():
6320 from zipimport
import zipimporter
6321 if hasattr(sys
, 'frozen'): # Running from PyInstaller
6322 path
= os
.path
.dirname(sys
.executable
)
6323 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
6324 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
6326 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
6327 return os
.path
.abspath(path
)
6330 def load_plugins(name
, suffix
, namespace
):
6333 plugins_spec
= importlib
.util
.spec_from_file_location(
6334 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
6335 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
6336 sys
.modules
[plugins_spec
.name
] = plugins
6337 plugins_spec
.loader
.exec_module(plugins
)
6338 for name
in dir(plugins
):
6339 if name
in namespace
:
6341 if not name
.endswith(suffix
):
6343 klass
= getattr(plugins
, name
)
6344 classes
[name
] = namespace
[name
] = klass
6345 except FileNotFoundError
:
6351 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
6352 casesense
=True, is_user_input
=False, traverse_string
=False):
6353 ''' Traverse nested list/dict/tuple
6354 @param path_list A list of paths which are checked one by one.
6355 Each path is a list of keys where each key is a string,
6356 a function, a tuple of strings or "...".
6357 When a fuction is given, it takes the key as argument and
6358 returns whether the key matches or not. When a tuple is given,
6359 all the keys given in the tuple are traversed, and
6360 "..." traverses all the keys in the object
6361 @param default Default value to return
6362 @param expected_type Only accept final value of this type (Can also be any callable)
6363 @param get_all Return all the values obtained from a path or only the first one
6364 @param casesense Whether to consider dictionary keys as case sensitive
6365 @param is_user_input Whether the keys are generated from user input. If True,
6366 strings are converted to int/slice if necessary
6367 @param traverse_string Whether to traverse inside strings. If True, any
6368 non-compatible object will also be converted into a string
6372 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
6373 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
6375 def _traverse_obj(obj
, path
, _current_depth
=0):
6379 path
= tuple(variadic(path
))
6380 for i
, key
in enumerate(path
):
6381 if isinstance(key
, (list, tuple)):
6382 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
6385 obj
= (obj
.values() if isinstance(obj
, dict)
6386 else obj
if isinstance(obj
, (list, tuple, LazyList
))
6387 else str(obj
) if traverse_string
else [])
6389 depth
= max(depth
, _current_depth
)
6390 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
6392 if isinstance(obj
, (list, tuple, LazyList
)):
6393 obj
= enumerate(obj
)
6394 elif isinstance(obj
, dict):
6397 if not traverse_string
:
6401 depth
= max(depth
, _current_depth
)
6402 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
6403 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
6404 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
6405 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
6408 key
= (int_or_none(key
) if ':' not in key
6409 else slice(*map(int_or_none
, key
.split(':'))))
6410 if key
== slice(None):
6411 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
6412 if not isinstance(key
, (int, slice)):
6414 if not isinstance(obj
, (list, tuple, LazyList
)):
6415 if not traverse_string
:
6424 if isinstance(expected_type
, type):
6425 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
6426 elif expected_type
is not None:
6427 type_test
= expected_type
6429 type_test
= lambda val
: val
6431 for path
in path_list
:
6433 val
= _traverse_obj(obj
, path
)
6436 for _
in range(depth
- 1):
6437 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
6438 val
= [v
for v
in map(type_test
, val
) if v
is not None]
6440 return val
if get_all
else val
[0]
6442 val
= type_test(val
)
6448 def traverse_dict(dictn
, keys
, casesense
=True):
6449 ''' For backward compatibility. Do not use '''
6450 return traverse_obj(dictn
, keys
, casesense
=casesense
,
6451 is_user_input
=True, traverse_string
=True)
6454 def variadic(x
, allowed_types
=(str, bytes)):
6455 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
6458 # create a JSON Web Signature (jws) with HS256 algorithm
6459 # the resulting format is in JWS Compact Serialization
6460 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
6461 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
6462 def jwt_encode_hs256(payload_data
, key
, headers
={}):
6468 header_data
.update(headers
)
6469 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
6470 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
6471 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
6472 signature_b64
= base64
.b64encode(h
.digest())
6473 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
6477 def supports_terminal_sequences(stream
):
6478 if compat_os_name
== 'nt':
6479 if get_windows_version() < (10, 0, 10586):
6481 elif not os
.getenv('TERM'):
6484 return stream
.isatty()
6485 except BaseException
:
6489 TERMINAL_SEQUENCES
= {
6492 'ERASE_LINE': '\x1b[K',
6493 'RED': '\033[0;31m',
6494 'YELLOW': '\033[0;33m',
6495 'BLUE': '\033[0;34m',
6496 'RESET_STYLE': '\033[0m',